My 11th 2023 update

components/utils/keywords/CMakeLists.txt

@@ -0,0 +1,2 @@
+add_library(keywords keywords_rule.cc single_keyword.cc data_keyword.cc pcre_keyword.cc length_keyword.cc byte_extract_keyword.cc compare_keyword.cc jump_keyword.cc stateop_keyword.cc no_match_keyword.cc)
+add_subdirectory(keywords_ut)

components/utils/keywords/byte_extract_keyword.cc

@@ -0,0 +1,319 @@
+#include "single_keyword.h"
+#include "output.h"
+#include "debug.h"
+
+#include <map>
+#include <strings.h>
+#include "limits.h"
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class ByteExtractKeyword : public SingleKeyword
+{
+public:
+    explicit ByteExtractKeyword(const vector<KeywordAttr> &attr, VariablesMapping &vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState *prev) const override;
+
+private:
+    enum class BaseId
+    {
+        BIN,
+        HEX = 16,
+        DEC = 10,
+        OCT = 8
+    };
+
+    void
+    setOffset(const KeywordAttr &attr, const VariablesMapping &vars)
+    {
+        offset.setAttr(attr, vars, "byte_extract");
+    }
+
+    void
+    setRelative(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_relative.setAttr(attr, "byte_extract");
+    }
+
+    void
+    setLittleEndian(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_little_end.setAttr(attr, "byte_extract");
+    }
+
+    void
+    setDataType(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        if (data_type != BaseId::BIN) {
+            throw KeywordError("Double definition of the data type in the 'byte_extract' keyword");
+        }
+
+        auto &vec = attr.getParams();
+        if (vec.size() != 2) throw KeywordError("Malformed data type in the 'byte_extract' keyword");
+
+        if (vec[1] == "hex") {
+            data_type = BaseId::HEX;
+        } else if (vec[1] == "dec") {
+            data_type = BaseId::DEC;
+        } else if (vec[1] == "oct") {
+            data_type = BaseId::OCT;
+        } else {
+            throw KeywordError("Unknown data type in the 'byte_extract' keyword: " + vec[1]);
+        }
+    }
+
+    void
+    setContext(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        ctx.setAttr(attr, "byte_extract");
+    }
+
+    void
+    setAlign(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        if (align != 1) throw KeywordError("Double definition of the 'align' in the 'byte_extract' keyword");
+        auto &vec = attr.getParams();
+        if (vec.size() != 2) throw KeywordError("Malformed 'align' in the 'byte_extract' keyword");
+
+        if (vec[1] == "2") {
+            align = 2;
+        } else if (vec[1] == "4") {
+            align = 4;
+        } else {
+            throw KeywordError("Unknown 'align' in the 'byte_extract' keyword: " + vec[1]);
+        }
+    }
+
+    bool
+    isConstant() const
+    {
+        return !is_relative && bytes.isConstant() && offset.isConstant();
+    }
+
+    pair<uint, uint> getStartOffsetAndLength(uint buf_size, const I_KeywordRuntimeState *prev) const;
+    uint applyAlignment(uint value) const;
+    Maybe<uint> readValue(uint start, uint length, const Buffer &buf) const;
+    Maybe<uint> readStringValue(uint start, uint length, const Buffer &buf) const;
+
+    NumericAttr bytes;
+    uint        var_id;
+    NumericAttr offset;
+    BoolAttr    is_relative;
+    BoolAttr    is_little_end;
+    BaseId      data_type = BaseId::BIN;
+    int         align = 1;
+    CtxAttr     ctx;
+
+    static const map<string, void(ByteExtractKeyword::*)(const KeywordAttr &, const VariablesMapping &)> setops;
+};
+
+const map<string, void(ByteExtractKeyword::*)(const KeywordAttr &, const VariablesMapping &)>
+ByteExtractKeyword::setops = {
+    { "offset",             &ByteExtractKeyword::setOffset       },
+    { "relative",           &ByteExtractKeyword::setRelative     },
+    { "little_endian",      &ByteExtractKeyword::setLittleEndian },
+    { "string",             &ByteExtractKeyword::setDataType     },
+    { "part",               &ByteExtractKeyword::setContext      },
+    { "align",              &ByteExtractKeyword::setAlign        }
+};
+
+ByteExtractKeyword::ByteExtractKeyword(const vector<KeywordAttr> &attrs, VariablesMapping &vars)
+        :
+    offset()
+{
+    //two requied attributes - number of bytes and var name
+    if (attrs.size() < 2) throw KeywordError("Invalid number of attributes in the 'byte_extract' keyword");
+
+    //parisng first attribute (Required) - number of bytes
+    auto &bytes_param = attrs[0].getParams();
+    if (bytes_param.size() != 1) {
+        throw KeywordError("More than one element in the 'bytes' in the 'byte_extract' keyword");
+    }
+    bytes.setAttr("bytes", bytes_param[0], vars, "byte_extract", static_cast<uint>(BaseId::DEC), true);
+    if (bytes.isConstant() && bytes.evalAttr(nullptr) == 0) {
+        throw KeywordError("Number of bytes is zero in the 'byte_extract' keyword");
+    }
+
+    //parisng second attribute (Required) - variable name
+    auto &var_name_param = attrs[1].getParams();
+    if (var_name_param.size() != 1) {
+        throw KeywordError("More than one element in the variable name in the 'byte_extract' keyword");
+    }
+    const string &var_name = var_name_param[0];
+    auto curr = setops.find(var_name);
+    if (curr != setops.end()) {
+        throw KeywordError("'" + var_name + "' cannot be the variable name in the 'byte_extract' keyword");
+    }
+    if (isdigit(var_name[0]) || var_name[0] == '-') {
+        throw KeywordError("Malformed variable name in the 'byte_extract' keyword");
+    }
+
+    var_id = vars.addNewVariable(var_name);
+
+    //parsing the other optional attributes
+    for (uint i = 2; i < attrs.size(); i++) {
+        auto curr = setops.find(attrs[i].getAttrName());
+        if (curr == setops.end()) {
+            throw KeywordError("Unknown attribute '" + attrs[i].getAttrName() + "' in the 'byte_extract' keyword");
+        }
+        auto set_func = curr->second;
+        (this->*set_func)(attrs[i], vars);
+    }
+
+    if (data_type == BaseId::BIN) {
+        if (!bytes.isConstant()) {
+            throw KeywordError("Data type is binary, but the 'bytes' is not constant in the 'byte_extract' keyword");
+        }
+        int num_bytes = bytes.evalAttr(nullptr);
+        if (num_bytes != 1 && num_bytes != 2 && num_bytes != 4) {
+            throw KeywordError("Data type is binary, but the 'bytes' is not constant in the 'byte_extract' keyword");
+        }
+        if (is_little_end && num_bytes == 1) {
+            throw KeywordError(
+                "Little endian is set, "
+                "but the number of bytes is invalid in the 'byte_extract' keyword"
+            );
+        }
+        if (align != 1) {
+            throw KeywordError("The 'align' is set and data type is binary in the 'byte_extract' keyword");
+        }
+    } else {
+        if (is_little_end) {
+            throw KeywordError("Little endian is set, but the data type is not binary in the 'byte_extract' keyword");
+        }
+    }
+}
+
+static uint
+addOffset(uint offset, int add)
+{
+    if (add < 0 && offset < static_cast<uint>(-add)) {
+        dbgWarning(D_KEYWORD)
+            << "The offset was set to 0 "
+            << "due to an attempt to jump before the beginning of the buffer in the 'jump' keyword";
+        return 0;
+    }
+    return offset + add;
+}
+
+pair<uint, uint>
+ByteExtractKeyword::getStartOffsetAndLength(uint buf_size, const I_KeywordRuntimeState *prev) const
+{
+    uint relative_offset = is_relative ? prev->getOffset(ctx) : 0;
+    int offset_attr = offset.evalAttr(prev);
+    uint start_offset = addOffset(relative_offset, offset_attr);
+
+    if (start_offset >= buf_size) return make_pair(0, 0);
+
+    uint length = buf_size - start_offset;
+    return make_pair(start_offset, length);
+}
+
+Maybe<uint>
+ByteExtractKeyword::readValue(uint start, uint length, const Buffer &buf) const
+{
+    if (data_type != BaseId::BIN) return readStringValue(start, length, buf);
+
+    uint res = 0;
+    for (uint i = 0; i < length; i++) {
+        uint ch = buf[start + i];
+        if (is_little_end) {
+            ch <<= 8*i;
+            res += ch;
+        } else {
+            res <<= 8;
+            res += ch;
+        }
+    }
+    return res;
+}
+
+Maybe<uint>
+ByteExtractKeyword::readStringValue(uint start, uint length, const Buffer &buf) const
+{
+    const u_char *data = buf.getPtr(start, length).unpack(); // start and length were checked outside of the function
+    string val_str(reinterpret_cast<const char *>(data), length);
+
+    uint base = static_cast<uint>(data_type);
+    try {
+        size_t idx;
+        auto res = stoul(val_str, &idx, base);
+        if (idx != val_str.length()) throw invalid_argument("");
+        if (res > INT_MAX) {
+            throw out_of_range("");
+        }
+        return res;
+    }
+    catch (invalid_argument &) {
+        return genError("Unable to convert the \"" + val_str + "\" to a number due to an invalid argument");
+    }
+    catch (out_of_range &) {
+        return genError(
+            "Unable to convert the \""
+            + val_str
+            + "\" to a number. The maximum is: "
+            + to_string(INT_MAX)
+        );
+    }
+}
+
+uint
+ByteExtractKeyword::applyAlignment(uint value) const
+{
+    int reminder = value % align;
+    if (reminder != 0) {
+        value += (align - reminder);
+    }
+    return value;
+}
+
+MatchStatus
+ByteExtractKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    auto part = Singleton::Consume<I_Environment>::by<KeywordComp>()->get<Buffer>(static_cast<string>(ctx));
+
+    if (!part.ok()) return MatchStatus::NoMatchFinal;
+
+    uint bytes_to_extr = bytes.evalAttr(prev);
+    if (bytes_to_extr == 0) {
+        dbgDebug(D_KEYWORD) << "Number of bytes is zero in the 'byte_extract' keyword";
+        return MatchStatus::NoMatch;  //the case of constant number of bytes was checked during compilation
+    }
+    
+    uint start_offset, length_to_end;
+    tie(start_offset, length_to_end) = getStartOffsetAndLength((*part).size(), prev);
+
+    uint offset_after_extracted_bytes = applyAlignment(start_offset + bytes_to_extr);
+
+    if (length_to_end == 0 || offset_after_extracted_bytes > (*part).size()) {
+        dbgDebug(D_KEYWORD)
+            << "Offset after the number of bytes to extract exceeds the buffer size in the 'byte_extract' keyword";
+        return isConstant() ? MatchStatus::NoMatchFinal : MatchStatus::NoMatch;
+    }
+
+    auto res = readValue(start_offset, bytes_to_extr, *part);
+    if (!res.ok()) {
+        dbgDebug(D_KEYWORD) << "Trying to store an invalid value in the 'byte_extract' keyword: " + res.getErr();
+        return isConstant() ? MatchStatus::NoMatchFinal : MatchStatus::NoMatch;
+    }
+
+    uint extracted_val = res.unpack();
+
+    if (extracted_val > INT_MAX) {
+        dbgDebug(D_KEYWORD) << "Value exceeds the maximum in the 'byte_extract' keyword";
+        return isConstant() ? MatchStatus::NoMatchFinal : MatchStatus::NoMatch;
+    }
+
+    //add variable and move offset after number of extracted bytes
+    VariableRuntimeState new_var(prev, var_id, extracted_val);
+    OffsetRuntimeState new_offset(&new_var, ctx, offset_after_extracted_bytes);
+    return runNext(&new_offset);
+}
+
+unique_ptr<SingleKeyword>
+genByteExtractKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<ByteExtractKeyword>(attr, known_vars);
+}

components/utils/keywords/compare_keyword.cc

@@ -0,0 +1,70 @@
+#include "single_keyword.h"
+#include "output.h"
+#include "debug.h"
+
+#include <map>
+#include <strings.h>
+#include "limits.h"
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class CompareKeyword : public SingleKeyword
+{
+public:
+    explicit CompareKeyword(const vector<KeywordAttr> &attr, const VariablesMapping &vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState* prev) const override;
+
+private:
+    bool
+    isConstant() const
+    {
+        return first_val.isConstant() && second_val.isConstant();
+    }
+
+    NumericAttr    first_val;
+    NumericAttr    second_val;
+    ComparisonAttr comparison;
+};
+
+CompareKeyword::CompareKeyword(const vector<KeywordAttr> &attrs, const VariablesMapping &vars)
+{
+    if (attrs.size() != 3) throw KeywordError("Invalid number of attributes in the 'compare' keyword");
+    
+    auto &first_val_param = attrs[0].getParams();
+    if (first_val_param.size() != 1) {
+        throw KeywordError("More than one element in the first value in the 'compare' keyword");
+    }
+    first_val.setAttr("first_val", first_val_param[0], vars, "compare");
+
+    auto &comparison_param = attrs[1].getParams();
+    if (comparison_param.size() != 1) {
+        throw KeywordError("More than one element in the comparison operator in the 'compare' keyword");
+    }
+    comparison.setAttr(comparison_param[0], "compare");
+
+    auto &second_val_param = attrs[2].getParams();
+    if (second_val_param.size() != 1) {
+        throw KeywordError("More than one element in the second value in the 'compare' keyword");
+    }
+    second_val.setAttr("second_val", second_val_param[0], vars, "compare");
+}
+
+MatchStatus
+CompareKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    int keyword_first_val = first_val.evalAttr(prev);
+    int keyword_second_val = second_val.evalAttr(prev);
+
+    if (comparison(keyword_first_val, keyword_second_val)) return runNext(prev);
+
+    // If there was no matches and the keyword is effected by other keywords, then we know that the rule won't match
+    return isConstant() ? MatchStatus::NoMatchFinal : MatchStatus::NoMatch;
+}
+
+unique_ptr<SingleKeyword>
+genCompareKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<CompareKeyword>(attr, known_vars);
+}

components/utils/keywords/data_keyword.cc

@@ -0,0 +1,409 @@
+#include "single_keyword.h"
+#include "output.h"
+#include "debug.h"
+
+#include <map>
+#include <strings.h>
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class DataKeyword : public SingleKeyword
+{
+public:
+    explicit DataKeyword(const vector<KeywordAttr> &attr, const VariablesMapping &vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState* prev) const override;
+
+private:
+    void
+    setOffset(const KeywordAttr &attr, const VariablesMapping &vars)
+    {
+        offset.setAttr(attr, vars, "data");
+    }
+
+    void
+    setDepth(const KeywordAttr &attr, const VariablesMapping &vars)
+    {
+        depth.setAttr(attr, vars, "data");
+    }
+
+    void
+    setCaret(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_caret.setAttr(attr, "data");
+    }
+
+    void
+    setRelative(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_relative.setAttr(attr, "data");
+    }
+
+    void
+    setCaseInsensitive(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_case_insensitive.setAttr(attr, "data");
+    }
+
+    void
+    setContext(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        ctx.setAttr(attr, "data");
+    }
+
+    void parseString(const string &str);
+
+    void
+    addChar(char ch)
+    {
+        pattern.push_back(static_cast<unsigned char>(ch));
+    }
+
+    void calcTables();
+
+    pair<uint, uint> getStartAndEndOffsets(uint buf_size, const I_KeywordRuntimeState *prev) const;
+    uint bytesMatched(const Buffer&, uint) const;
+
+    uint
+    moveOnMatch() const
+    {
+        return pattern.size();
+    }
+
+    uint
+    moveOnNoMatch(uint offset_from_end, unsigned char first_unmatched_byte) const
+    {
+        dbgAssert(shift.size() > offset_from_end) << "Shift table of the 'data' keyword is shorter than the offset";
+
+        uint skip_size;
+        if (skip[first_unmatched_byte]>offset_from_end) {
+            skip_size = skip[first_unmatched_byte]-offset_from_end;
+        } else {
+            skip_size = 1;
+        }
+
+        return max(shift[offset_from_end], skip_size);
+    }
+
+    bool
+    isConstant() const
+    {
+        return !is_relative && offset.isConstant() && depth.isConstant();
+    }
+
+    vector<unsigned char> pattern;
+    uint                  skip[256];
+    vector<uint>          shift;
+
+    NumericAttr           offset;
+    NumericAttr           depth;
+    BoolAttr              is_negative;
+    BoolAttr              is_caret;
+    BoolAttr              is_relative;
+    BoolAttr              is_case_insensitive;
+    CtxAttr               ctx;
+
+    static const map<string, void(DataKeyword::*)(const KeywordAttr &, const VariablesMapping &)> setops;
+};
+
+const map<string, void(DataKeyword::*)(const KeywordAttr &, const VariablesMapping &)> DataKeyword::setops = {
+    { "relative", &DataKeyword::setRelative        },
+    { "offset",   &DataKeyword::setOffset          },
+    { "depth",    &DataKeyword::setDepth           },
+    { "caret",    &DataKeyword::setCaret           },
+    { "nocase",   &DataKeyword::setCaseInsensitive },
+    { "part",     &DataKeyword::setContext         }
+};
+
+DataKeyword::DataKeyword(const vector<KeywordAttr> &attrs, const VariablesMapping &vars)
+        :
+    offset(),
+    depth()
+{
+    auto &pattern_param = attrs[0].getParams();
+
+    if (pattern_param.size() != 1) throw KeywordError("More than one element in the 'data' keyword pattern");
+    const string &string_pattern = pattern_param[0];
+
+    if (string_pattern.length() == 0) throw KeywordError("No input for the 'data' keyword");
+
+    uint start = 0;
+    if (string_pattern[0] == '!') {
+        is_negative.setAttr("data", "negative");
+        start++;
+    }
+    if (string_pattern[start] != '"') throw KeywordError("The data pattern does not begin with '\"'");
+
+    uint end = string_pattern.length()-1;
+    if (string_pattern[end] != '"') throw KeywordError("The data pattern does not end with '\"'");
+
+    if (start+1 >= end) throw KeywordError("No input for the 'data' keyword");
+
+    parseString(string_pattern.substr(start+1, end-(start+1)));
+
+    for (uint i = 1; i<attrs.size(); i++) {
+        auto curr = setops.find(attrs[i].getAttrName());
+        if (curr == setops.end()) {
+            throw KeywordError("Unknown attribute '" + attrs[i].getAttrName() + "' in the 'data' keyword");
+        }
+        auto set_func = curr->second;
+        (this->*set_func)(attrs[i], vars);
+    }
+
+    calcTables();
+}
+
+void
+DataKeyword::calcTables()
+{
+    if (is_case_insensitive) {
+        for (auto &ch : pattern) {
+            if (isupper(ch)) {
+                ch = tolower(ch);
+            }
+        }
+    }
+
+    // Initialize skip table - when we meet a charecter that isn't in the pattern, we skip the whole pattern
+    for (auto &ch_skip : skip) {
+        ch_skip = pattern.size();
+    }
+
+    // Go over the charecters in the pattern.
+    // We can skip from a charecter to the end of the pattern.
+    // If a charecter appear more than once, the latest occurence take precedent.
+    for (uint index = 0; index<pattern.size(); index++) {
+        unsigned char ch = pattern[index];
+
+        uint dist_to_end = pattern.size()-(index+1);
+        if (is_case_insensitive && islower(ch)) {
+            skip[toupper(ch)] = dist_to_end;
+        }
+        skip[ch] = dist_to_end;
+    }
+
+    // Initialize the shift table.
+    shift.resize(pattern.size(), 0);
+
+    uint end_offset = pattern.size()-1;
+    // Go over all the suffixes (from the empty to the pattern-1)
+    for (size_t suffix_len = 0; suffix_len<pattern.size(); suffix_len++) {
+        // Find the smallest shift, so when shifting the suffix left:
+        // 1. All chars overlapping between pattern and shifted suffix match.
+        // 2. If the character before the shifted suffix overlaps the pattern, it doesn't match.
+        // pattern = "hellololo", suff=2 (must match "[^o]lo"), shift=4 ("hel(lo)lolo")
+        // pattern = "olo" suff=2 (must match "[^o]lo"), shift=2 ("(.o)lo")
+        //           characters before the patterns are considered wild.
+        for (uint shift_offset = 1; shift_offset<=pattern.size(); shift_offset++) {
+            // Verify that in the current offset matches the suffix
+            size_t num_of_overlapping_char;
+            unsigned char *suffix_start_ptr;
+            unsigned char *shifted_suffix_start_ptr;
+            if (shift_offset+suffix_len <= pattern.size()) {
+                // Shifted suffix doesn't exceed the pattern. Compare the whole suffix.
+                num_of_overlapping_char  = suffix_len;
+                suffix_start_ptr         = pattern.data() + pattern.size() - suffix_len;
+                shifted_suffix_start_ptr = suffix_start_ptr - shift_offset;
+            } else {
+                // Shifted suffix exceeds the pattern. Compare only the overlaping charecters.
+                num_of_overlapping_char  = pattern.size() - shift_offset;
+                suffix_start_ptr         = pattern.data() + shift_offset;
+                shifted_suffix_start_ptr = pattern.data();
+            }
+
+            if (bcmp(suffix_start_ptr, shifted_suffix_start_ptr, num_of_overlapping_char) != 0) continue;
+
+            // Verify that what comes after the suffix doesn't match
+            if (shift_offset+suffix_len < pattern.size()) {
+                if (pattern[end_offset-suffix_len] == pattern[end_offset-(shift_offset+suffix_len)]) continue;
+            }
+
+            // Set the currect shift offset
+            shift[suffix_len] = shift_offset;
+            break;
+        }
+    }
+}
+
+void
+DataKeyword::parseString(const string &str)
+{
+    string hex;
+    bool hex_mode = false;
+    bool after_bslash = false;
+
+    for (auto ch : str) {
+        if (after_bslash) {
+            if (!isprint(ch)) {
+                throw KeywordError(
+                    "Illegal backslash character '" +
+                    dumpHexChar(ch) +
+                    "' in the pattern in the 'data' keyword"
+                );
+            }
+            addChar(ch);
+            after_bslash = false;
+            continue;
+        }
+
+        switch (ch) {
+        case '|': {
+            if (!hex_mode) {
+                hex = "";
+                hex_mode = true;
+            } else {
+                if (hex.size()>0) throw KeywordError("Stoping in the middle of hex string in the 'data' keyword");
+                hex_mode = false;
+            }
+            break;
+        }
+        case '\\': {
+            if (hex_mode) throw KeywordError("Backslash in hex string in the 'data' keyword");
+            after_bslash = true;
+            break;
+        }
+        case '"': {
+            throw KeywordError("Unescaped double quotation mark in the 'data' keyword");
+            break;
+        }
+        default:
+            if (hex_mode) {
+                if (!isxdigit(ch)) {
+                    if (ch != ' ') {
+                        throw KeywordError(
+                            "Illegal character '" +
+                            dumpHexChar(ch) +
+                            "' in the hex string in the 'data' keyword"
+                        );
+                    }
+                    if (hex.size()>0) {
+                        throw KeywordError("Space separating nibbles in the hex string in the 'data' keyword");
+                    }
+                    break;
+                }
+
+                hex += ch;
+
+                if (hex.size()>=2) {
+                    addChar(stol(hex, nullptr, 16));
+                    hex = "";
+                }
+            } else {
+                if (!isprint(ch)) {
+                    throw KeywordError(
+                        "Illegal character '" +
+                        dumpHexChar(ch) +
+                        "' in the pattern in the 'data' keyword"
+                    );
+                }
+                addChar(ch);
+            }
+        }
+    }
+
+    if ( hex_mode || after_bslash ) {
+        throw KeywordError("The 'data' keyword's pattern has ended in the middle of the parsing");
+    }
+}
+
+static uint
+addOffset(uint offset, int add)
+{
+    if (add<0 && offset<static_cast<uint>(-add)) return 0;
+    return offset + add;
+}
+
+pair<uint, uint>
+DataKeyword::getStartAndEndOffsets(uint buf_size, const I_KeywordRuntimeState *prev) const
+{
+    uint relative_offset = is_relative?prev->getOffset(ctx):0;
+    int offset_attr = offset.evalAttr(prev);
+    uint start_offset = addOffset(relative_offset, offset_attr);
+
+    if (depth.isSet()) {
+        uint depth_size = addOffset(start_offset, depth.evalAttr(prev));
+        buf_size = std::min(buf_size, depth_size);
+    }
+    if (is_caret) {
+        buf_size = std::min(buf_size, start_offset+static_cast<uint>(pattern.size()));
+    }
+
+    return make_pair(start_offset, buf_size);
+}
+
+uint
+DataKeyword::bytesMatched(const Buffer &buf, uint offset) const
+{
+    if (is_case_insensitive) {
+        for (uint i = 0;  i<pattern.size(); i++) {
+            if (pattern[pattern.size()-(i+1)] != tolower(buf[offset-(i+1)])) return i;
+        }
+    } else {
+        for (uint i = 0 ; i < pattern.size() ; i++ ) {
+            if (pattern[pattern.size()-(i+1)] != buf[offset-(i+1)] ) return i;
+        }
+    }
+    return pattern.size();
+}
+
+MatchStatus
+DataKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    dbgAssert(pattern.size()>0) << "Trying to run on an uninitialized keyword data";
+
+    dbgDebug(D_KEYWORD) << "Searching for " << dumpHex(pattern);
+
+    auto part = Singleton::Consume<I_Environment>::by<KeywordComp>()->get<Buffer>(static_cast<string>(ctx));
+    if (!part.ok()) {
+        if (is_negative) return runNext(prev);
+        return MatchStatus::NoMatchFinal;
+    }
+
+    const auto &buf = part.unpack();
+
+    dbgTrace(D_KEYWORD) << "Full buffer: " << dumpHex(buf);
+
+    uint offset, max_offset;
+
+    tie(offset, max_offset) = getStartAndEndOffsets(buf.size(), prev);
+    offset += pattern.size();
+
+    bool match_found = false;
+    while (offset<=max_offset) {
+        // Short circuit for the common, simple case where the last byte doesn't match
+        if (skip[buf[offset-1]]) {
+            offset += skip[buf[offset - 1]];
+            continue;
+        }
+
+        // Full search Boyer-Moore
+        uint match_size = bytesMatched(buf, offset);
+        if (match_size == pattern.size()) {
+            if (is_negative) {
+                return isConstant()?MatchStatus::NoMatchFinal:MatchStatus::NoMatch;
+            }
+            match_found = true;
+            OffsetRuntimeState new_offset(prev, ctx, offset);
+            auto next_keyword_result = runNext(&new_offset);
+            if (next_keyword_result!=MatchStatus::NoMatch) return next_keyword_result;
+            offset += moveOnMatch();
+        } else {
+            offset += moveOnNoMatch(match_size, buf[offset-(match_size+1)]);
+        }
+    }
+
+    // No matchs is a success for negative keywords
+    if (is_negative && !match_found) return runNext(prev);
+
+    // If there were no matchs and the keyword is an effected by other keywords, then we know that the rule won't match
+    if (isConstant() && !match_found) return MatchStatus::NoMatchFinal;
+
+    return MatchStatus::NoMatch;
+}
+
+unique_ptr<SingleKeyword>
+genDataKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<DataKeyword>(attr, known_vars);
+}

components/utils/keywords/jump_keyword.cc

@@ -0,0 +1,174 @@
+#include "single_keyword.h"
+#include "output.h"
+#include "debug.h"
+
+#include <map>
+#include <strings.h>
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class jumpKeyword : public SingleKeyword
+{
+public:
+    explicit jumpKeyword(const vector<KeywordAttr> &attr, const VariablesMapping &vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState* prev) const override;
+
+private:
+    enum class JumpFromId
+    {
+        RELATIVE,
+        FROM_BEGINNING,
+        FROM_END
+    };
+
+    void
+    setContext(const KeywordAttr &attr)
+    {
+        ctx.setAttr(attr, "byte_extract");
+    }
+
+    void
+    setAlign(const KeywordAttr &attr)
+    {
+        if (align != 1) throw KeywordError("Double definition of the 'align' in the 'jump' keyword");
+        auto &vec = attr.getParams();
+        if (vec.size() != 2) throw KeywordError("Malformed 'align' in the 'jump' keyword");
+
+        if (vec[1] == "2") {
+            align = 2;
+        } else if (vec[1] == "4") {
+            align = 4;
+        } else {
+            throw KeywordError("Unknown 'align' in the 'jump' keyword: " + vec[1]);
+        }
+    }
+    
+    bool
+    isConstant() const
+    {
+        return jumping_from != JumpFromId::RELATIVE && jumping_val.isConstant();
+    }
+
+    JumpFromId  jumping_from;
+    NumericAttr jumping_val;
+    int         align = 1;
+    CtxAttr     ctx;
+
+    static const map<string, void(jumpKeyword::*)(const KeywordAttr &)> setops;
+    uint getStartOffset(uint buf_size, const I_KeywordRuntimeState *prev) const;
+    uint applyAlignment(uint value) const;
+    uint addOffset(uint offset, int add) const;
+};
+
+const map<string, void(jumpKeyword::*)(const KeywordAttr &)> jumpKeyword::setops = {
+    { "part",            &jumpKeyword::setContext   },
+    { "align",           &jumpKeyword::setAlign     }
+};
+
+jumpKeyword::jumpKeyword(const vector<KeywordAttr> &attrs, const VariablesMapping &vars)
+{
+
+    //two requied attributes - jumping value and jumping from
+    if (attrs.size() < 2) throw KeywordError("Invalid number of attributes in the 'jump' keyword");
+
+    //parisng first attribute (Required) - jumping value
+    auto &jumping_val_param = attrs[0].getParams();
+    if (jumping_val_param.size() != 1) {
+        throw KeywordError("More than one element in the jumping value in the 'jump' keyword");
+    }
+    jumping_val.setAttr("jumping value", jumping_val_param[0], vars, "jump");
+
+    //parisng second attribute (Required) - jumping from
+    auto &jumping_from_param = attrs[1].getParams();
+    if (jumping_from_param.size() != 1) {
+        throw KeywordError("More than one element in the jumping 'from' parameter in the 'jump' keyword");
+    }
+    
+    if (jumping_from_param[0] == "from_beginning") {
+        jumping_from = JumpFromId::FROM_BEGINNING;
+    } else if (jumping_from_param[0] == "from_end") {
+        jumping_from = JumpFromId::FROM_END;
+    } else if (jumping_from_param[0] == "relative") {
+        jumping_from = JumpFromId::RELATIVE;
+    } else {
+        throw KeywordError("Unknown jumping 'from' parameter in the 'jump' keyword: " + jumping_from_param[0]);
+    }
+
+    //parisng optional attributes
+    for (uint i = 2; i < attrs.size(); i++) {
+        auto curr = setops.find(attrs[i].getAttrName());
+        if (curr == setops.end()) {
+            throw KeywordError("Unknown attribute " + attrs[i].getAttrName() + " in the 'jump' keyword");
+        }
+        auto set_func = curr->second;
+        (this->*set_func)(attrs[i]);
+    }
+}
+
+uint
+jumpKeyword::applyAlignment(uint value) const
+{
+    int reminder = value % align;
+    if (reminder != 0) {
+        value += (align - reminder);
+    }
+    return value;
+}
+
+uint
+jumpKeyword::addOffset(uint offset, int add) const
+{
+    if (add < 0 && offset < static_cast<uint>(-add)) {
+        dbgWarning(D_KEYWORD)
+            << "The offset was set to 0 "
+            << "due to an attempt to jump before the beginning of the buffer in the 'jump' keyword";
+        return 0;
+    }
+    return applyAlignment(offset + add);
+}
+
+uint
+jumpKeyword::getStartOffset(uint buf_size, const I_KeywordRuntimeState *prev) const
+{
+    switch (jumping_from) {
+        case JumpFromId::FROM_BEGINNING: {
+            return 0;
+        }
+        case JumpFromId::FROM_END: {
+            return buf_size;
+        }
+        case JumpFromId::RELATIVE: {
+            return prev->getOffset(ctx);
+        }
+    }
+    dbgAssert(false) << "Invalid jumping 'from' parameter";
+    return 0;
+}
+
+MatchStatus
+jumpKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    auto part = Singleton::Consume<I_Environment>::by<KeywordComp>()->get<Buffer>(static_cast<string>(ctx));
+
+    if (!part.ok()) return MatchStatus::NoMatchFinal;
+
+    uint start_offset = getStartOffset((*part).size(), prev);
+
+    uint offset_to_jump = addOffset(start_offset, jumping_val.evalAttr(prev));
+
+    if (offset_to_jump > (*part).size()) {
+        dbgDebug(D_KEYWORD) << "New offset exceeds the buffer size in the 'jump' keyword";
+        return isConstant() ? MatchStatus::NoMatchFinal : MatchStatus::NoMatch;
+    }
+
+    OffsetRuntimeState new_offset(prev, ctx, offset_to_jump);
+    return runNext(&new_offset);
+}
+
+unique_ptr<SingleKeyword>
+genJumpKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<jumpKeyword>(attr, known_vars);
+}

components/utils/keywords/keywords_rule.cc

@@ -0,0 +1,171 @@
+#include "keyword_comp.h"
+
+#include <vector>
+#include "sentinel_runtime_state.h"
+
+using namespace std;
+
+static const string whitespaces = " \t";
+
+static string
+getSubStrNoPadding(const string &str, uint start, uint end)
+{
+    auto r_start = str.find_first_not_of(whitespaces, start);
+    auto r_end = str.find_last_not_of(whitespaces, end-1);
+
+    if (r_end==string::npos || r_start==string::npos || r_start>r_end) {
+        throw KeywordError("Found an empty section in the '"+ str + "'");
+    }
+
+    return str.substr(r_start, r_end-r_start+1);
+}
+
+static vector<string>
+split(const string &str, const string &delim, uint start = 0)
+{
+    vector<string> res;
+    uint part_start = start;
+    bool escape = false;
+    bool in_string = false;
+
+    for (uint index = start; index<str.size(); index++) {
+        if (escape) {
+            escape = false;
+            continue;
+        }
+
+        switch (str[index]) {
+            case '\\': {
+                escape = true;
+                break;
+            }
+            case '"': {
+                in_string = !in_string;
+                break;
+            }
+            default:
+                if (!in_string && delim.find(str[index])!=string::npos) {
+                    res.push_back(getSubStrNoPadding(str, part_start, index));
+                    part_start = index+1;
+                }
+        }
+    }
+
+    if (escape||in_string) throw KeywordError("Split has ended in the middle of the parsing");
+
+    if (str.find_first_not_of(whitespaces, part_start)!=string::npos) {
+        res.push_back(getSubStrNoPadding(str, part_start, str.size()));
+    }
+
+    return res;
+}
+
+KeywordAttr::KeywordAttr(const string &str) : params(split(str, whitespaces))
+{
+}
+
+KeywordParsed::KeywordParsed(string const &keyword) {
+    auto index = keyword.find_first_of(':');
+    if (index!=string::npos) {
+        for (auto &str : split(keyword, ",", index+1)) {
+            attr.push_back(KeywordAttr(str));
+        }
+    } else {
+        index = keyword.size();
+    }
+
+    name = getSubStrNoPadding(keyword, 0, index);
+    if (name.find_first_of(whitespaces)!=string::npos) {
+        throw KeywordError("'" + name + "' - cannot be a keyword name");
+    }
+}
+
+uint
+SentinelRuntimeState::getOffset(const std::string &) const
+{
+    return 0;
+}
+
+// LCOV_EXCL_START Reason: this function is tested in one_element_list_negative_test but marked as not covered.
+uint
+SentinelRuntimeState::getVariable(uint var_id) const
+{
+    dbgAssert(false) << "Could not find the variable ID: " << var_id;
+    return 0;
+}
+// LCOV_EXCL_STOP
+
+class SentinelKeyword : public SingleKeyword
+{
+public:
+    MatchStatus
+    isMatch() const
+    {
+        SentinelRuntimeState curr_state;
+        return runNext(&curr_state);
+    }
+
+private:
+    // LCOV_EXCL_START Reason: Unreachable function.
+    MatchStatus
+    isMatch(const I_KeywordRuntimeState *state) const override
+    {
+        return runNext(state);
+    }
+    // LCOV_EXCL_STOP
+};
+
+class KeywordComp::Impl : Singleton::Provide<I_KeywordsRule>::From<KeywordComp>
+{
+public:
+    Maybe<shared_ptr<VirtualRule>>
+    genRule(const string &rule)
+    {
+        shared_ptr<VirtualRule> res;
+        try {
+            res = KeywordsRuleImpl::genRule(rule);
+        } catch (const KeywordError &e) {
+            return genError(e.getErr());;
+        }
+        return move(res);
+    }
+
+private:
+    class KeywordsRuleImpl : public VirtualRule
+    {
+    public:
+        bool isMatch() const override { return start.isMatch() == MatchStatus::Match; }
+
+        static unique_ptr<KeywordsRuleImpl>
+        genRule(const string &rule)
+        {
+            auto res = make_unique<KeywordsRuleImpl>();
+
+            auto pos = rule.find_last_not_of(whitespaces);
+            if (pos==string::npos) {
+                // Empty rule
+                return res;
+            }
+
+            if (rule[pos]!=';') throw KeywordError(rule + " - end of text pass rule");
+
+            VariablesMapping known_vars;
+
+            auto key_vec = split(rule, ";");
+            for (auto &keyword : key_vec) {
+                res->start.appendKeyword(getKeywordByName(keyword, known_vars));
+            }
+
+            return res;
+        }
+
+    private:
+        SentinelKeyword start;
+    };
+};
+
+KeywordComp::KeywordComp() : Component("KeywordComp"), pimpl(make_unique<KeywordComp::Impl>()) {}
+
+KeywordComp::~KeywordComp() {}
+
+string I_KeywordsRule::keywords_tag = "keywords_rule_tag";

components/utils/keywords/keywords_ut/CMakeLists.txt

@@ -0,0 +1,5 @@
+add_unit_test(
+    keywords_ut
+    "keywords_ut.cc;single_keyword_ut.cc"
+    "keywords;pcre2-8;buffers;singleton;table;event_is;metric;-lboost_regex"
+)

components/utils/keywords/keywords_ut/keywords_ut.cc

@@ -0,0 +1,996 @@
+#include "keyword_comp.h"
+#include "environment.h"
+#include "mock/mock_table.h"
+#include "cptest.h"
+#include "mock/mock_time_get.h"
+#include "mock/mock_mainloop.h"
+#include "config.h"
+#include "config_component.h"
+
+using namespace std;
+
+class KeywordsRuleTest : public ::testing::Test
+{
+public:
+    void
+    appendBuffer(const string &id, const string &str)
+    {
+        buffers[id] += Buffer(str);
+    }
+
+    string
+    ruleCompileFail(const string &_rule)
+    {
+        auto rule = Singleton::Consume<I_KeywordsRule>::from(comp)->genRule(_rule);
+        EXPECT_FALSE(rule.ok()) << "Compile supposed to fail";
+        return rule.getErr();
+    }
+
+    bool
+    ruleRun(const string &_rule, const string &default_ctx = "default")
+    {
+        auto rule = Singleton::Consume<I_KeywordsRule>::from(comp)->genRule(_rule);
+        EXPECT_TRUE(rule.ok()) << "Compile not supposed to fail: " << rule.getErr();
+        ScopedContext ctx;
+        ctx.registerValue(I_KeywordsRule::getKeywordsRuleTag(), default_ctx);
+        for (auto &value : buffers) {
+            ctx.registerValue(value.first, value.second);
+        }
+        return (*rule)->isMatch();
+    }
+
+private:
+    KeywordComp comp;
+    ::testing::NiceMock<MockMainLoop> mock_mainloop;
+    ::testing::NiceMock<MockTimeGet> mock_timer;
+    Environment env;
+    map<string, Buffer> buffers;
+};
+
+TEST_F(KeywordsRuleTest, data_basic_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123456789");
+
+    EXPECT_TRUE(ruleRun("data: \"234\" , part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("data: \"234\";", "HTTP_RESPONSE_BODY"));
+    EXPECT_FALSE(ruleRun("data: \"75\", part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, data_relative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("data: \"567\", part HTTP_RESPONSE_BODY; data: \"234\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(
+        ruleRun("data: \"567\", part HTTP_RESPONSE_BODY; data: \"234\", part HTTP_RESPONSE_BODY, relative;")
+    );
+    EXPECT_TRUE(ruleRun("data: \"234\", part HTTP_RESPONSE_BODY; data: \"567\", part HTTP_RESPONSE_BODY, relative;"));
+}
+
+TEST_F(KeywordsRuleTest, data_depth_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("data: \"345\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("data: \"345\", depth 5, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("data: \"345\", depth 4, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, data_nocase_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "abcdefg");
+
+    EXPECT_TRUE(ruleRun("data: \"cde\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("data: \"CDE\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("data: \"CDE\", nocase, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, data_offset_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("data: \"345\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("data: \"345\", offset 2, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("data: \"345\", offset 3, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, data_caret_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("data: \"345\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("data: \"345\", part HTTP_RESPONSE_BODY, caret;"));
+    EXPECT_TRUE(ruleRun("data: \"345\", caret, part HTTP_RESPONSE_BODY, offset 2;"));
+}
+
+TEST_F(KeywordsRuleTest, data_negative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_FALSE(ruleRun("data: !\"345\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("data: !\"365\", part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, data_part_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+    appendBuffer("HTTP_REQUEST_BODY", "abcdefg");
+
+    EXPECT_TRUE(ruleRun("data: \"345\", part HTTP_RESPONSE_BODY; data: \"cde\", part HTTP_REQUEST_BODY;"));
+    EXPECT_FALSE(ruleRun("data: \"345\", part HTTP_RESPONSE_BODY; data: \"cde\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("data: \"345\", part HTTP_REQUEST_BODY; data: \"cde\", part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, pcre_basic_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("pcre: \"/5.7/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/5..7/\", part HTTP_RESPONSE_BODY;"));
+}
+
+
+TEST_F(KeywordsRuleTest, pcre_relative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("pcre: \"/5.7/\", part HTTP_RESPONSE_BODY; pcre: \"/2.4/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/5.7/\", part HTTP_RESPONSE_BODY; pcre: \"/2.4/R\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(
+        ruleRun("pcre: \"/5.7/\", part HTTP_RESPONSE_BODY; pcre: \"/2.4/\", relative, part HTTP_RESPONSE_BODY;")
+    );
+    EXPECT_TRUE(ruleRun("pcre: \"/2.4/\", part HTTP_RESPONSE_BODY; pcre: \"/5.7/R\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(
+        ruleRun("pcre: \"/2.4/\", part HTTP_RESPONSE_BODY; pcre: \"/5.7/\", relative, part HTTP_RESPONSE_BODY;")
+    );
+}
+
+TEST_F(KeywordsRuleTest, pcre_depth_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("pcre: \"/3.5/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("pcre: \"/3.5/\", depth 5, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/3.5/\", depth 4, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, pcre_nocase_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "abcdefg");
+
+    EXPECT_TRUE(ruleRun("pcre: \"/c.e/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/C.E/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("pcre: \"/C.E/i\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("pcre: \"/C.E/\", nocase, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, pcre_offset_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("pcre: \"/3.5/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("pcre: \"/3.5/\", offset 2, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/3.5/\", offset 300, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, pcre_part_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+    appendBuffer("HTTP_REQUEST_BODY", "abcdefg");
+
+    EXPECT_TRUE(ruleRun("pcre: \"/3.5/\", part HTTP_RESPONSE_BODY; pcre: \"/c.e/\", part HTTP_REQUEST_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/3.5/\", part HTTP_RESPONSE_BODY; pcre: \"/c.e/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("pcre: \"/3.5/\", part HTTP_REQUEST_BODY; pcre: \"/c.e/\", part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, pcre_negative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_FALSE(ruleRun("pcre: !\"/3.5/\", part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("pcre: !\"/3..5/\", part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, compare_comparison_test) {
+    EXPECT_TRUE(ruleRun("compare: 0, =, 0;"));
+    EXPECT_TRUE(ruleRun("compare: -1, =, -1;"));
+    EXPECT_FALSE(ruleRun("compare: 0, =, 1;"));
+    EXPECT_FALSE(ruleRun("compare: -1, =, -2;"));
+    EXPECT_FALSE(ruleRun("compare: 1, =, -1;"));
+    EXPECT_FALSE(ruleRun("compare: -1, =, 1;"));
+    EXPECT_TRUE(ruleRun("compare: 2, !=, 3;"));
+    EXPECT_TRUE(ruleRun("compare: 2, <=, 3;"));
+    EXPECT_TRUE(ruleRun("compare: 2, <, 3;"));
+    EXPECT_FALSE(ruleRun("compare: 2, >, 3;"));
+    EXPECT_FALSE(ruleRun("compare: 2, >=, 3;"));
+    EXPECT_TRUE(ruleRun("compare: -2, !=, -3;"));
+    EXPECT_TRUE(ruleRun("compare: -2, >=, -3;"));
+    EXPECT_TRUE(ruleRun("compare: -2, >, -3;"));
+    EXPECT_FALSE(ruleRun("compare: -2, <, -3;"));
+    EXPECT_FALSE(ruleRun("compare: -2, <=, -3;"));
+    EXPECT_TRUE(ruleRun("compare: -2, !=, 3;"));
+    EXPECT_TRUE(ruleRun("compare: -2, <=, 3;"));
+    EXPECT_TRUE(ruleRun("compare: -2, <, 3;"));
+    EXPECT_FALSE(ruleRun("compare: -2, >, 3;"));
+    EXPECT_FALSE(ruleRun("compare: -2, >=, 3;"));
+    EXPECT_TRUE(ruleRun("compare: 2, !=, -3;"));
+    EXPECT_TRUE(ruleRun("compare: 2, >=, -3;"));
+    EXPECT_TRUE(ruleRun("compare: 2, >, -3;"));
+    EXPECT_FALSE(ruleRun("compare: 2, <, -3;"));
+    EXPECT_FALSE(ruleRun("compare: 2, <=, -3;"));
+}
+
+TEST_F(KeywordsRuleTest, compare_compile_fail_test) {
+    EXPECT_EQ(ruleCompileFail("compare: 0;"), "Invalid number of attributes in the 'compare' keyword");
+    EXPECT_EQ(ruleCompileFail("compare: 0, =;"), "Invalid number of attributes in the 'compare' keyword");
+    EXPECT_EQ(ruleCompileFail("compare: 0, =, 0, 0;"), "Invalid number of attributes in the 'compare' keyword");
+    EXPECT_EQ(
+        ruleCompileFail("compare: 0 1, =, 0;"),
+        "More than one element in the first value in the 'compare' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("compare: 0, = =, 0;"),
+        "More than one element in the comparison operator in the 'compare' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("compare: 0, =, 0 1;"),
+        "More than one element in the second value in the 'compare' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("compare: 0, ==, 0;"),
+        "Unknown comparison operator in the 'compare' keyword: Could not find the operator: =="
+    );
+}
+
+TEST_F(KeywordsRuleTest, length_basic_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123456789");
+    appendBuffer("HTTP_REQUEST_BODY", "");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "length: length_var, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 9;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "length: length_var, part HTTP_REQUEST_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "length: length_var, part HTTP_REQUEST_BODY;"
+            "compare: length_var, =, 1;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, length_part_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("length: length_var, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("length: length_var, part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, length_relative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123456789");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "data: \"234\", part HTTP_RESPONSE_BODY;"
+            "length: relative_length_var, part HTTP_RESPONSE_BODY, relative;"
+            "compare: relative_length_var, =, 5;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "data: \"234\", part HTTP_RESPONSE_BODY;"
+            "length: relative_length_var, part HTTP_RESPONSE_BODY;"
+            "compare: relative_length_var, =, 5;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "data: \"89\", part HTTP_RESPONSE_BODY;"
+            "length: zero_length_var, part HTTP_RESPONSE_BODY, relative;"
+            "compare: zero_length_var, =, 0;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "data: \"89\", part HTTP_RESPONSE_BODY;"
+            "length: zero_length_var, part HTTP_RESPONSE_BODY;"
+            "compare: zero_length_var, =, 0;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, length_compare_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123");
+    EXPECT_FALSE(ruleRun("length: 6, min, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("length: 6, exact, part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("length: 6, max, part HTTP_RESPONSE_BODY;"));
+
+    appendBuffer("HTTP_RESPONSE_BODY", "456");
+    EXPECT_TRUE(ruleRun("length: 6, min, part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("length: 6, exact, part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(ruleRun("length: 6, max, part HTTP_RESPONSE_BODY;"));
+
+    appendBuffer("HTTP_RESPONSE_BODY", "789");
+    EXPECT_TRUE(ruleRun("length: 6, min, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("length: 6, exact, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("length: 6, max, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, length_compile_fail_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123456789");
+
+    EXPECT_EQ(
+        ruleCompileFail("length: two_elem 2, part HTTP_RESPONSE_BODY;"),
+        "More than one element in the variable name in the 'length' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("length: relative, part HTTP_RESPONSE_BODY;"),
+        "The 'relative' cannot be the variable name in the 'length' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("length: part, part HTTP_RESPONSE_BODY;"),
+        "The 'part' cannot be the variable name in the 'length' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("length: -minus, part HTTP_RESPONSE_BODY;"),
+        "Malformed variable name in the 'length' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("length: 1digit, part HTTP_RESPONSE_BODY;"),
+        "Malformed variable name in the 'length' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("length: bad_attr, partt HTTP_RESPONSE_BODY;"),
+        "Unknown attribute 'partt' in the 'length' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("length:;"),
+        "Invalid number of attributes in the 'length' keyword"
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_dec_string_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, dec_var, string dec, part HTTP_RESPONSE_BODY;"
+            "data: \"234\", offset dec_var, part HTTP_RESPONSE_BODY;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 1, dec_var, string dec, part HTTP_RESPONSE_BODY;"
+            "data: \"123\", offset dec_var, part HTTP_RESPONSE_BODY;"
+        )
+    );
+
+    appendBuffer("HTTP_REQUEST_BODY", "A");
+
+    EXPECT_FALSE(ruleRun("byte_extract: 1, bad_dec_var, string dec, part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_hex_string_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "A123");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, hex_var, string hex, part HTTP_RESPONSE_BODY;"
+            "compare: hex_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 2, hex_var, string hex, part HTTP_RESPONSE_BODY;"
+            "compare: hex_var, =, 161;"
+        )
+    );
+
+    appendBuffer("HTTP_REQUEST_BODY", "10G");
+
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 2, hex_var, string hex, part HTTP_REQUEST_BODY;"
+            "compare: hex_var, =, 10;"
+        )
+    );
+    EXPECT_FALSE(ruleRun("byte_extract: 3, bad_hex_var, string oct, part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_oct_string_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "13ABC");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 2, oct_var, string oct, part HTTP_RESPONSE_BODY;"
+            "compare: oct_var, =, 11;"
+        )
+    );
+
+    appendBuffer("HTTP_REQUEST_BODY", "118");
+
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 2, oct_var, string oct, part HTTP_REQUEST_BODY;"
+            "compare: oct_var, =, 13;"
+        )
+    );
+    EXPECT_FALSE(ruleRun("byte_extract: 3, bad_oct_var, string oct, part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_binary_data_test) {
+    string one_byte_binary_data = {10};
+    appendBuffer("HTTP_RESPONSE_BODY", one_byte_binary_data);
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, binary_data_var, part HTTP_RESPONSE_BODY;"
+            "compare: binary_data_var, =, 10;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 1, dec_data_var, offset 2, string dec, part HTTP_RESPONSE_BODY;"
+            "compare: dec_data_var, =, 10;"
+        )
+    );
+
+    string two_bytes_binary_data = {1, 0, 0};
+    appendBuffer("HTTP_REQUEST_BODY", two_bytes_binary_data);
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 2, binary_data_var, part HTTP_REQUEST_BODY;"
+            "compare: binary_data_var , =, 256;"
+        )
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 3, not1/2/4, part HTTP_REQUEST_BODY;"),
+        "Data type is binary, but the 'bytes' is not constant in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail(
+            "byte_extract: 1, no_constant, part HTTP_REQUEST_BODY;"
+            "byte_extract: no_constant, var, part HTTP_REQUEST_BODY;"
+        ),
+        "Data type is binary, but the 'bytes' is not constant in the 'byte_extract' keyword"
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_bad_num_of_bytes_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "0");
+
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 0, zero_bytes_var, string dec, part HTTP_RESPONSE_BODY;"),
+        "Number of bytes is zero in the 'byte_extract' keyword"
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 1, one_byte_var, string dec, part HTTP_RESPONSE_BODY;"
+            "byte_extract: one_byte_var, zero_bytes_var, string dec, part HTTP_RESPONSE_BODY;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_part_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123");
+    EXPECT_TRUE(ruleRun("byte_extract: 1, part_var, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("byte_extract: 1, part_var, part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_offset_test) {
+    appendBuffer("HTTP_REQUEST_BODY", "1A23456789hello");
+
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 1, hex_var, offset 1, string hex, part HTTP_REQUEST_BODY; "
+            "data: \"9hell\", offset hex_var, part HTTP_REQUEST_BODY;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, hex_var, offset 1, string hex, part HTTP_REQUEST_BODY;"
+            "data: \"hell\", offset hex_var, part HTTP_REQUEST_BODY;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 1, dec_var, offset -1, string dec, part HTTP_REQUEST_BODY;"
+            "data: \"1A2\", offset dec_var, part HTTP_REQUEST_BODY;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, dec_var, offset -1, string dec, part HTTP_REQUEST_BODY;"
+            "data: \"A2\", offset dec_var, part HTTP_REQUEST_BODY;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_relative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "123456789");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "data: \"12\", part HTTP_RESPONSE_BODY;"
+            "byte_extract: 1, relative_var, relative, string dec, part HTTP_RESPONSE_BODY;"
+            "compare: relative_var, =, 3;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "data: \"12\", part HTTP_RESPONSE_BODY;"
+            "byte_extract: 1, non_relative_var, string dec, part HTTP_RESPONSE_BODY;"
+            "compare: non_relative_var, =, 3;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "data: \"89\", part HTTP_RESPONSE_BODY;"
+            "byte_extract: 1, relative_var, string dec, relative, part HTTP_RESPONSE_BODY;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_endianness_test) {
+    string little_end_test_str = {8, 0, 0};
+    appendBuffer("HTTP_RESPONSE_BODY", little_end_test_str);
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 2, lit_end_var, little_endian, part HTTP_RESPONSE_BODY;"
+            "compare: lit_end_var, =, 8;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 2, big_end_var, part HTTP_RESPONSE_BODY;"
+            "compare: big_end_var, =, 8;"
+        )
+    );
+
+    little_end_test_str[1] = 0;
+    little_end_test_str[2] = 1;
+    appendBuffer("HTTP_REQUEST_BODY", little_end_test_str);
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 2, lit_end_with_offset_var,"
+            "offset 1, little_endian, part HTTP_REQUEST_BODY;"
+            "compare: lit_end_with_offset_var, =, 256;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: 2, big_end_with_offset_var, offset 1, part HTTP_REQUEST_BODY;"
+            "compare: big_end_with_offset_var, =, 256;"
+        )
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, var, little_endian, part HTTP_REQUEST_BODY;"),
+        "Little endian is set, but the number of bytes is invalid in the 'byte_extract' keyword"
+    );
+
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 2, no_binary, little_endian, string dec, part HTTP_REQUEST_BODY;"),
+        "Little endian is set, but the data type is not binary in the 'byte_extract' keyword"
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_align_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, align2_var, align 2, string dec, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 2;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, align4_var, align 4, string dec, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, align2_var, offset 3, align 2, string dec, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, align4_var, offset 3, align 4, string dec, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+
+    appendBuffer("HTTP_REQUEST_BODY", "123");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: 1, align2_var, offset 1, align 2, string dec, part HTTP_REQUEST_BODY;"
+            "length: length_var, relative, part HTTP_REQUEST_BODY;"
+            "compare: length_var, =, 1;"
+        )
+    );
+    EXPECT_FALSE(ruleRun("byte_extract: 1, align4_var, align 4, string dec, part HTTP_REQUEST_BODY;"));
+    EXPECT_FALSE(ruleRun("byte_extract: 1, align2_var, offset 2, align 2, string dec, part HTTP_REQUEST_BODY;"));
+
+    string binary_data_str = { 1 };
+    appendBuffer("HTTP_REQUEST_BODY", binary_data_str);
+
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, align_binary_var, align 2, part HTTP_REQUEST_BODY;"),
+        "The 'align' is set and data type is binary in the 'byte_extract' keyword"
+    );
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_overflow_test) {
+    string overflow_dec_data_str = to_string((uint)INT_MAX + 1);
+    appendBuffer("HTTP_RESPONSE_BODY", overflow_dec_data_str);
+
+    EXPECT_FALSE(
+        ruleRun(
+            "byte_extract: " + to_string(overflow_dec_data_str.length()) + ","
+            "overflow_var, string dec, part HTTP_RESPONSE_BODY;"
+        )
+    );
+
+    string max_value_dec_data_str = to_string(INT_MAX);
+    appendBuffer("HTTP_REQUEST_BODY", max_value_dec_data_str);
+
+    EXPECT_TRUE(
+        ruleRun(
+            "byte_extract: " + to_string(max_value_dec_data_str.length()) + ","
+            "max_var, string dec, part HTTP_REQUEST_BODY;"
+            "compare: max_var, =, " + max_value_dec_data_str + ";"
+        )
+    );
+
+    string overflow_binary_data_str = { 0x7f, 0x7f, 0x7f, 0x7f, 0 };
+    appendBuffer("HTTP_REQUEST_HEADERS", overflow_binary_data_str);
+
+    EXPECT_FALSE(ruleRun("byte_extract: 5 ,overflow_num_var, string dec, part HTTP_REQUEST_HEADERS;"));
+}
+
+TEST_F(KeywordsRuleTest, byte_extract_compile_fail_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1 2, dec_var, string dec, part HTTP_RESPONSE_BODY;"),
+        "More than one element in the 'bytes' in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, dec_var 1, string dec, part HTTP_RESPONSE_BODY;"),
+        "More than one element in the variable name in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, align, string dec, part HTTP_RESPONSE_BODY;"),
+        "'align' cannot be the variable name in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, -1, string dec, part HTTP_RESPONSE_BODY;"),
+        "Malformed variable name in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_data_type, string dechex, part HTTP_RESPONSE_BODY;"),
+        "Unknown data type in the 'byte_extract' keyword: dechex"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, 1var, string dec, part HTTP_RESPONSE_BODY;"),
+        "Malformed variable name in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_align, align 3, part HTTP_RESPONSE_BODY;"),
+        "Unknown 'align' in the 'byte_extract' keyword: 3"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_constant, offset 0x;"),
+        "Malformed constant '0x' in the 'offset' in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(ruleCompileFail("byte_extract: 1;"), "Invalid number of attributes in the 'byte_extract' keyword");
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_attr, offset;"),
+        "Malformed offset' in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_attr, string hex dec;"),
+        "Malformed data type in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_attr, ofset 5;"),
+        "Unknown attribute 'ofset' in the 'byte_extract' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("byte_extract: 1, bad_align, align 2 4;"),
+        "Malformed 'align' in the 'byte_extract' keyword"
+    );
+}
+
+TEST_F(KeywordsRuleTest, jump_from_beginning_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 9;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 9;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: -1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "length: length_var, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 10, from_beginning, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+    EXPECT_FALSE(ruleRun("jump: 11, from_beginning, part HTTP_RESPONSE_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, jump_relative_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "data: \"1\", part HTTP_RESPONSE_BODY;"
+            "jump: 1, relative, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 8;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "data: \"1\", part HTTP_RESPONSE_BODY;"
+            "jump: 1, relative, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 9;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "data: \"1\", part HTTP_RESPONSE_BODY;"
+            "jump: -2, relative, part HTTP_RESPONSE_BODY;"
+            "length: length_var, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "data: \"1\", part HTTP_RESPONSE_BODY;"
+            "jump: 9, relative, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "data: \"1\", part HTTP_RESPONSE_BODY;"
+            "jump: 10, relative, part HTTP_RESPONSE_BODY;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, jump_from_end_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_FALSE(ruleRun("jump: 1, from_end, part HTTP_RESPONSE_BODY;"));
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: -1, from_end, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 1;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: -10, from_end, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: -11, from_end, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, combined_jumps_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "jump: -1, from_end, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 1;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"
+            "jump: -1, relative, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: -1, from_end, part HTTP_RESPONSE_BODY;"
+            "jump: 1, relative, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 0;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: -1, from_end, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 1;"
+        )
+    );
+}
+
+
+TEST_F(KeywordsRuleTest, jump_alignment_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 1, from_beginning, align 2, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 8;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 1, from_beginning, align 4, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 6;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 3, from_beginning, part HTTP_RESPONSE_BODY;"
+            "jump: 2, relative, align 2, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 4;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "jump: 3, from_beginning, part HTTP_RESPONSE_BODY;"
+            "jump: 2, relative, align 2, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 7;"
+        )
+    );
+    EXPECT_FALSE(
+        ruleRun(
+            "jump: 3, from_beginning, part HTTP_RESPONSE_BODY;"
+            "jump: 2, relative, align 4, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 3;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 2, from_beginning, align 2, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 8;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 4, from_beginning, align 4, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 6;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 0, from_beginning, align 2, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+    EXPECT_TRUE(
+        ruleRun(
+            "jump: 0, from_beginning, align 4, part HTTP_RESPONSE_BODY;"
+            "length: length_var, relative, part HTTP_RESPONSE_BODY;"
+            "compare: length_var, =, 10;"
+        )
+    );
+}
+
+TEST_F(KeywordsRuleTest, jump_part_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_TRUE(ruleRun("jump: 1, from_beginning, part HTTP_RESPONSE_BODY;"));
+    EXPECT_FALSE(ruleRun("jump: 1, from_beginning, part HTTP_REQUEST_BODY;"));
+}
+
+TEST_F(KeywordsRuleTest, jump_compile_fail_test) {
+    appendBuffer("HTTP_RESPONSE_BODY", "1234567890");
+
+    EXPECT_EQ(ruleCompileFail("jump: 1;"), "Invalid number of attributes in the 'jump' keyword");
+    EXPECT_EQ(
+        ruleCompileFail("jump: 2 1, from_beginning;"),
+        "More than one element in the jumping value in the 'jump' keyword"
+    );
+    EXPECT_EQ(
+        ruleCompileFail("jump: 2, from_relative;"),
+        "Unknown jumping 'from' parameter in the 'jump' keyword: from_relative"
+    );
+    EXPECT_EQ(ruleCompileFail("jump: 2, relative, align 3;"), "Unknown 'align' in the 'jump' keyword: 3");
+    EXPECT_EQ(ruleCompileFail("jump: 2, relative, align 1;"), "Unknown 'align' in the 'jump' keyword: 1");
+    EXPECT_EQ(ruleCompileFail("jump: 2, relative, align2 2;"), "Unknown attribute align2 in the 'jump' keyword");
+    EXPECT_EQ(ruleCompileFail("jump: 2, relative, align 2 4;"), "Malformed 'align' in the 'jump' keyword");
+    EXPECT_EQ(
+        ruleCompileFail("jump: 2, from_beginning relative;"),
+        "More than one element in the jumping 'from' parameter in the 'jump' keyword"
+    );
+}
+
+TEST_F(KeywordsRuleTest, stateop)
+{
+    using testing::_;
+
+    ConfigComponent conf;
+    testing::StrictMock<MockTable> table;
+
+    std::unique_ptr<TableOpaqueBase> opq;
+    TableOpaqueBase *opq_ptr;
+    bool has_stage = false;
+    EXPECT_CALL(table, createStateRValueRemoved(_, _))
+        .WillOnce(testing::DoAll(
+            testing::Invoke(
+                [&] (const type_index &, std::unique_ptr<TableOpaqueBase> &other)
+                {
+                    opq = std::move(other);
+                    opq_ptr = opq.get();
+                    has_stage = true;
+                }
+            ),
+            testing::Return(true)
+        ));
+    EXPECT_CALL(table, getState(_)).WillRepeatedly(testing::ReturnPointee(&opq_ptr));
+    EXPECT_CALL(table, hasState(_)).WillRepeatedly(testing::ReturnPointee(&has_stage));
+
+    EXPECT_FALSE(ruleRun("stateop: state sss, isset;"));
+
+    EXPECT_TRUE(ruleRun("stateop: state sss, unset;"));
+    EXPECT_FALSE(ruleRun("stateop: state sss, isset;"));
+
+    EXPECT_TRUE(ruleRun("stateop: state sss, set;"));
+    EXPECT_TRUE(ruleRun("stateop: state sss, isset;"));
+    EXPECT_FALSE(ruleRun("stateop: state dd, isset;"));
+
+    EXPECT_TRUE(ruleRun("stateop: state sss, unset;"));
+    EXPECT_FALSE(ruleRun("stateop: state sss, isset;"));
+}
+
+TEST_F(KeywordsRuleTest, no_match)
+{
+    EXPECT_FALSE(ruleRun("no_match;"));
+}

components/utils/keywords/keywords_ut/single_keyword_ut.cc

@@ -0,0 +1,171 @@
+#include "../sentinel_runtime_state.h"
+#include "../single_keyword.h"
+#include "cptest.h"
+#include <list>
+
+using namespace std;
+using namespace testing;
+
+#define FIRST_VARIABLE_ID 1
+#define FIRST_VARIABLE_VAL 2u
+#define SECOND_VARIABLE_ID 3
+#define SECOND_VARIABLE_VAL 4u
+#define THIRD_VARIABLE_ID 5
+#define THIRD_VARIABLE_VAL 6u
+
+#define FIRST_OFFSET 4u
+#define SECOND_OFFSET 5u
+#define THIRD_OFFSET 6u
+
+const static unsigned int zero = 0;
+
+class I_KeywordRuntimeStateTest : public Test
+{
+public:
+    //  Constructs SentinelKeyword as head because it is the only I_KeywordRuntimeState implementation
+    //  that does not hold I_KeywordRuntimeState *prev
+    I_KeywordRuntimeStateTest() : list_head(&sentinel) {}
+
+    uint
+    getOffset(const string &id)
+    {
+        return list_head->getOffset(id);
+    }
+
+    uint
+    getVariable(uint requested_var_id)
+    {
+        return list_head->getVariable(requested_var_id);
+    }
+
+    void
+    addOffsetState(const string &_ctx, uint _offset)
+    {
+        offset_list.push_front(make_unique<OffsetRuntimeState>(list_head, _ctx, _offset));
+        list_head = offset_list.front().get();
+    }
+
+    void
+    addVariableState(uint _var_id, uint _val)
+    {
+        variable_list.push_front(make_unique<VariableRuntimeState>(list_head, _var_id, _val));
+        list_head = variable_list.front().get();
+    }
+
+private:
+    list<unique_ptr<VariableRuntimeState>> variable_list;
+    list<unique_ptr<OffsetRuntimeState>> offset_list;
+    SentinelRuntimeState sentinel;
+    I_KeywordRuntimeState *list_head;
+};
+
+TEST_F(I_KeywordRuntimeStateTest, one_element_list_positive_test) {
+    EXPECT_EQ(getOffset("HTTP_METHOD"), zero);
+    EXPECT_EQ(getOffset("HTTP_REQ_COOKIE"), zero);
+    EXPECT_EQ(getOffset("HTTP_REQUEST_HEADERS"), zero);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, one_variable_state_list_positive_test) {
+    addVariableState(FIRST_VARIABLE_ID, FIRST_VARIABLE_VAL);
+    EXPECT_EQ(getOffset("HTTP_COMPLETE_URL_ENCODED"), zero);
+    EXPECT_EQ(getOffset("HTTP_METHOD"), zero);
+    EXPECT_EQ(getVariable(FIRST_VARIABLE_ID), FIRST_VARIABLE_VAL);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, one_offset_state_list_positive_test) {
+    EXPECT_EQ(getOffset("HTTP_REQUEST_HEADERS"), zero);
+    addOffsetState("HTTP_REQUEST_HEADERS", FIRST_OFFSET);
+    EXPECT_EQ(getOffset("HTTP_REQUEST_HEADERS"), FIRST_OFFSET);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, one_element_list_negative_test) {
+    cptestPrepareToDie();
+    EXPECT_DEATH(getVariable(FIRST_OFFSET), "");
+    EXPECT_DEATH(getVariable(SECOND_OFFSET), "");
+    EXPECT_DEATH(getVariable(THIRD_OFFSET), "");
+}
+
+TEST_F(I_KeywordRuntimeStateTest, variable_runtime_state_list_positive_test) {
+    //  Notice that variables ids and values are different
+    addVariableState(FIRST_VARIABLE_ID, FIRST_VARIABLE_VAL);
+    addVariableState(SECOND_VARIABLE_ID, SECOND_VARIABLE_VAL);
+    addVariableState(THIRD_VARIABLE_ID, THIRD_VARIABLE_VAL);
+
+    EXPECT_EQ(getOffset("HTTP_METHOD"), zero);
+    EXPECT_EQ(getOffset("HTTP_COMPLETE_URL_ENCODED"), zero);
+    EXPECT_EQ(getOffset("HTTP_REQ_COOKIE"), zero);
+    EXPECT_EQ(getOffset("HTTP_REQUEST_HEADERS"), zero);
+    EXPECT_EQ(getOffset("HTTP_REQUEST_BODY"), zero);
+
+    EXPECT_EQ(getVariable(FIRST_VARIABLE_ID), FIRST_VARIABLE_VAL);
+    EXPECT_EQ(getVariable(SECOND_VARIABLE_ID), SECOND_VARIABLE_VAL);
+    EXPECT_EQ(getVariable(THIRD_VARIABLE_ID), THIRD_VARIABLE_VAL);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, OffsetRuntimeState_list_negative_test) {
+    addOffsetState("HTTP_COMPLETE_URL_ENCODED", FIRST_OFFSET);
+    addOffsetState("HTTP_REQ_COOKIE", SECOND_OFFSET);
+    addOffsetState("HTTP_METHOD", THIRD_OFFSET);
+
+    cptestPrepareToDie();
+    EXPECT_DEATH(getVariable(FIRST_OFFSET), "");
+    EXPECT_DEATH(getVariable(SECOND_OFFSET), "");
+    EXPECT_DEATH(getVariable(THIRD_OFFSET), "");
+}
+
+TEST_F(I_KeywordRuntimeStateTest, offset_runtime_state_list_positive_test) {
+    EXPECT_EQ(getOffset("HTTP_REQUEST_HEADERS"), zero);
+    EXPECT_EQ(getOffset("HTTP_REQ_COOKIE"), zero);
+
+    addOffsetState("HTTP_REQUEST_HEADERS", FIRST_OFFSET);
+    addOffsetState("HTTP_REQ_COOKIE", SECOND_OFFSET);
+
+    EXPECT_EQ(getOffset("HTTP_REQUEST_HEADERS"), FIRST_OFFSET);
+    EXPECT_EQ(getOffset("HTTP_REQ_COOKIE"), SECOND_OFFSET);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, mixed_types_list_positive_test) {
+    EXPECT_EQ(getOffset("HTTP_COMPLETE_URL_ENCODED"), zero);
+    EXPECT_EQ(getOffset("HTTP_METHOD"), zero);
+
+    addOffsetState("HTTP_COMPLETE_URL_ENCODED", FIRST_OFFSET);
+    addVariableState(SECOND_VARIABLE_ID, SECOND_VARIABLE_VAL);
+    addOffsetState("HTTP_METHOD", THIRD_OFFSET);
+
+    EXPECT_EQ(getOffset("HTTP_COMPLETE_URL_ENCODED"), FIRST_OFFSET);
+    EXPECT_EQ(getOffset("HTTP_METHOD"), THIRD_OFFSET);
+
+    EXPECT_EQ(getVariable(SECOND_VARIABLE_ID), SECOND_VARIABLE_VAL);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, mixed_types_list_negative_test) {
+    addOffsetState("HTTP_COMPLETE_URL_ENCODED", FIRST_OFFSET);
+    addVariableState(SECOND_VARIABLE_ID, SECOND_VARIABLE_VAL);
+    addOffsetState("HTTP_METHOD", THIRD_OFFSET);
+
+    cptestPrepareToDie();
+    EXPECT_DEATH(getVariable(FIRST_OFFSET), "");
+    EXPECT_DEATH(getVariable(THIRD_OFFSET), "");
+}
+
+TEST_F(I_KeywordRuntimeStateTest, mixed_types_list_offset_shadowing_test) {
+    addOffsetState("HTTP_COMPLETE_URL_ENCODED", FIRST_OFFSET);
+
+    EXPECT_EQ(getOffset("HTTP_COMPLETE_URL_ENCODED"), FIRST_OFFSET);
+
+    addVariableState(SECOND_VARIABLE_ID, SECOND_VARIABLE_VAL);
+    addOffsetState("HTTP_COMPLETE_URL_ENCODED", THIRD_OFFSET);
+
+    EXPECT_EQ(getOffset("HTTP_COMPLETE_URL_ENCODED"), THIRD_OFFSET);
+}
+
+TEST_F(I_KeywordRuntimeStateTest, mixed_types_list_variable_shadowing_test) {
+    addVariableState(FIRST_VARIABLE_ID, FIRST_VARIABLE_VAL);
+
+    EXPECT_EQ(getVariable(FIRST_VARIABLE_ID), FIRST_VARIABLE_VAL);
+
+    addOffsetState("HTTP_COMPLETE_URL_ENCODED", SECOND_OFFSET);
+    addVariableState(FIRST_VARIABLE_ID, THIRD_VARIABLE_VAL);
+
+    EXPECT_EQ(getVariable(FIRST_VARIABLE_ID), THIRD_VARIABLE_VAL);
+}

components/utils/keywords/length_keyword.cc

@@ -0,0 +1,162 @@
+#include "single_keyword.h"
+#include "output.h"
+#include "debug.h"
+#include "flags.h"
+
+#include <map>
+#include <strings.h>
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class LengthKeyword : public SingleKeyword
+{
+public:
+    explicit LengthKeyword(const vector<KeywordAttr> &attr, VariablesMapping &vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState* prev) const override;
+
+private:
+    enum class Mode { EXACT, MIN, MAX, COUNT };
+    using ModeFlags = Flags<Mode>;
+
+    void
+    setRelative(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_relative.setAttr(attr, "length");
+    }
+
+    void
+    setExact(const KeywordAttr &, const VariablesMapping &)
+    {
+        if (!mode.empty()) throw KeywordError("Redefining 'length' keyword operation");
+        mode.setFlag(Mode::EXACT);
+    }
+
+    void
+    setMin(const KeywordAttr &, const VariablesMapping &)
+    {
+        if (!mode.empty()) throw KeywordError("Redefining 'length' keyword operation");
+        mode.setFlag(Mode::MIN);
+    }
+
+
+    void
+    setMax(const KeywordAttr &, const VariablesMapping &)
+    {
+        if (!mode.empty()) throw KeywordError("Redefining 'length' keyword operation");
+        mode.setFlag(Mode::MAX);
+    }
+
+
+    void
+    setContext(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        ctx.setAttr(attr, "length");
+    }
+
+    bool
+    isConstant() const
+    {
+        return !is_relative && compare_size.isConstant();
+    }
+
+    BoolAttr    is_relative;
+    ModeFlags   mode;
+    CtxAttr     ctx;
+    uint        var_id;
+    NumericAttr compare_size;
+
+    static const map<string, void(LengthKeyword::*)(const KeywordAttr &, const VariablesMapping &)> setops;
+};
+
+const map<string, void(LengthKeyword::*)(const KeywordAttr &, const VariablesMapping &)> LengthKeyword::setops = {
+    { "relative",   &LengthKeyword::setRelative },
+    { "exact",      &LengthKeyword::setExact    },
+    { "min",        &LengthKeyword::setMin      },
+    { "max",        &LengthKeyword::setMax      },
+    { "part",       &LengthKeyword::setContext  }
+};
+
+LengthKeyword::LengthKeyword(const vector<KeywordAttr> &attrs, VariablesMapping &vars)
+{
+    if (attrs.size() == 0) throw KeywordError("Invalid number of attributes in the 'length' keyword");
+    
+    //parisng first attribute (Required) - variable name
+    auto &var_name_param = attrs[0].getParams();
+
+    if (var_name_param.size() != 1) {
+        throw KeywordError("More than one element in the variable name in the 'length' keyword");
+    }
+    
+    const string &string_var_name = var_name_param[0];
+
+    if (string_var_name == "relative") {
+        throw KeywordError("The 'relative' cannot be the variable name in the 'length' keyword");
+    }
+    if (string_var_name == "part") {
+        throw KeywordError("The 'part' cannot be the variable name in the 'length' keyword");
+    }
+    if (string_var_name == "exact") {
+        throw KeywordError("The 'exact' cannot be the variable name in the 'length' keyword");
+    }
+    if (string_var_name == "min") {
+        throw KeywordError("The 'min' cannot be the variable name in the 'length' keyword");
+    }
+    if (string_var_name == "max") {
+        throw KeywordError("The 'max' cannot be the variable name in the 'length' keyword");
+    }
+
+    //parsing the other optional attributes
+    for (uint i = 1; i < attrs.size(); i++) {
+        auto curr = setops.find(attrs[i].getAttrName());
+        if (curr == setops.end()) {
+            throw KeywordError("Unknown attribute '" + attrs[i].getAttrName() + "' in the 'length' keyword");
+        }
+        auto set_func = curr->second;
+        (this->*set_func)(attrs[i], vars);
+    }
+
+    if (mode.empty()) {
+        if (isdigit(string_var_name[0]) || string_var_name[0] == '-') {
+            throw KeywordError("Malformed variable name in the 'length' keyword");
+        }
+
+        var_id = vars.addNewVariable(string_var_name);
+    } else {
+        compare_size.setAttr("length value", string_var_name, vars, "length", 10, true);
+    }
+}
+
+MatchStatus
+LengthKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    auto part = Singleton::Consume<I_Environment>::by<KeywordComp>()->get<Buffer>(static_cast<string>(ctx));
+
+    if (!part.ok()) return MatchStatus::NoMatchFinal;
+
+    uint offset = is_relative ? prev->getOffset(ctx) : 0;
+    uint size = (*part).size();
+
+    if (offset <= size) {
+        if (mode.isSet(Mode::EXACT)) {
+            if (size - offset == static_cast<uint>(compare_size.evalAttr(prev))) return runNext(prev);
+        } else if (mode.isSet(Mode::MIN)) {
+            if (size - offset >= static_cast<uint>(compare_size.evalAttr(prev))) return runNext(prev);
+        } else if (mode.isSet(Mode::MAX)) {
+            if (size - offset <= static_cast<uint>(compare_size.evalAttr(prev))) return runNext(prev);
+        } else {
+            VariableRuntimeState new_length_var(prev, var_id, size-offset);
+            return runNext(&new_length_var);
+        }
+    }
+
+    // If there was no matches and the keyword is effected by other keywords, then we know that the rule won't match
+    return isConstant() ? MatchStatus::NoMatchFinal : MatchStatus::NoMatch;
+}
+
+unique_ptr<SingleKeyword>
+genLengthKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<LengthKeyword>(attr, known_vars);
+}

components/utils/keywords/no_match_keyword.cc

@@ -0,0 +1,29 @@
+#include "single_keyword.h"
+#include "table_opaque.h"
+#include "debug.h"
+
+#include <map>
+#include <strings.h>
+
+#include "cereal/types/set.hpp"
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class NoMatchKeyword : public SingleKeyword
+{
+public:
+    explicit NoMatchKeyword(const vector<KeywordAttr> &attr, VariablesMapping &)
+    {
+        if (!attr.empty()) throw KeywordError("The 'no_match' keyword doesn't take attributes");
+    }
+
+    MatchStatus isMatch(const I_KeywordRuntimeState *) const override { return MatchStatus::NoMatchFinal; }
+};
+
+unique_ptr<SingleKeyword>
+genNoMatchKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<NoMatchKeyword>(attr, known_vars);
+}

components/utils/keywords/pcre_keyword.cc

@@ -0,0 +1,372 @@
+#include "single_keyword.h"
+
+#include <algorithm>
+
+#define PCRE2_CODE_UNIT_WIDTH 8
+#include <pcre2.h>
+
+#include "output.h"
+#include "debug.h"
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class PCREKeyword : public SingleKeyword
+{
+public:
+    explicit PCREKeyword(const vector<KeywordAttr> &attr, const VariablesMapping &known_vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState *prev) const override;
+
+private:
+    void
+    setOffset(const KeywordAttr &attr, const VariablesMapping &vars)
+    {
+        offset.setAttr(attr, vars, "pcre");
+    }
+
+    void
+    setDepth(const KeywordAttr &attr, const VariablesMapping &vars)
+    {
+        depth.setAttr(attr, vars, "pcre");
+    }
+
+    void
+    setRelative(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_relative.setAttr(attr, "pcre");
+    }
+
+    void
+    setCaseInsensitive(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        is_case_insensitive.setAttr(attr, "pcre");
+    }
+
+    void
+    setContext(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        ctx.setAttr(attr, "pcre");
+    }
+
+    string parseString(const string &str);
+    pair<string, string> findExprInStr(const string &str, size_t start, size_t end);
+    void parseOptions(const string &str);
+    void compilePCRE(const string &str);
+
+    pair<uint, uint> getStartOffsetAndLength(uint buf_size, const I_KeywordRuntimeState *prev) const;
+
+    bool
+    isConstant() const
+    {
+        return !is_relative && offset.isConstant() && depth.isConstant();
+    }
+
+    class PCREDelete
+    {
+    public:
+        void
+        operator()(pcre2_code *ptr)
+        {
+            pcre2_code_free(ptr);
+        }
+    };
+    unique_ptr<pcre2_code, PCREDelete> pcre_machine;
+
+    class PCREResultDelete
+    {
+    public:
+        void
+        operator()(pcre2_match_data *ptr)
+        {
+            pcre2_match_data_free(ptr);
+        }
+    };
+    unique_ptr<pcre2_match_data, PCREResultDelete> pcre_result;
+
+    NumericAttr offset;
+    NumericAttr depth;
+    BoolAttr is_negative;
+    BoolAttr is_relative;
+
+    BoolAttr is_case_insensitive;
+    BoolAttr is_multiline;
+    BoolAttr is_dotall;
+    BoolAttr is_extended;
+    BoolAttr is_dollar_endonly;
+    BoolAttr is_anchor;
+    BoolAttr is_ungreedy;
+
+    CtxAttr ctx;
+
+    string pcre_expr;
+
+    static const map<string, void(PCREKeyword::*)(const KeywordAttr &, const VariablesMapping &)> setops;
+};
+
+const map<string, void(PCREKeyword::*)(const KeywordAttr&, const VariablesMapping&)> PCREKeyword::setops = {
+    { "relative", &PCREKeyword::setRelative        },
+    { "offset",   &PCREKeyword::setOffset          },
+    { "depth",    &PCREKeyword::setDepth           },
+    { "nocase",   &PCREKeyword::setCaseInsensitive },
+    { "part",     &PCREKeyword::setContext         },
+};
+
+PCREKeyword::PCREKeyword(const vector<KeywordAttr> &attrs, const VariablesMapping &known_vars)
+        :
+    offset(),
+    depth()
+{
+    auto &expr_param = attrs[0].getParams();
+    if (expr_param.size() != 1) throw KeywordError("More than one element in the 'pcre' keyword pattern");
+    auto expr = parseString(expr_param[0]);
+    dbgDebug(D_KEYWORD) << "Creating a new 'pcre' expression: " << expr;
+
+    for (uint i = 1; i<attrs.size(); i++) {
+        auto curr = setops.find(attrs[i].getAttrName());
+        if (curr == setops.end()) {
+            throw KeywordError("Unknown attribute '" + attrs[i].getAttrName() + "' in the 'pcre' keyword");
+        }
+        auto set_func = curr->second;
+        (this->*set_func)(attrs[i], known_vars);
+    }
+
+    compilePCRE(expr);
+}
+
+string
+PCREKeyword::parseString(const string &str)
+{
+    size_t start_offset = 0, end_offset = str.size();
+
+    if (start_offset<end_offset && str[start_offset]=='!') {
+        is_negative.setAttr("pcre", "negative");
+        start_offset++;
+    }
+
+    if (start_offset+1>=end_offset || str[start_offset]!='"' || str[end_offset-1]!='"') {
+        throw KeywordError("The 'pcre' expression should be enclosed in quotation marks");
+    }
+    start_offset++;
+    end_offset--;
+
+    string expr, options;
+    tie(expr, options) = findExprInStr(str, start_offset, end_offset);
+
+    parseOptions(options);
+
+    return expr;
+}
+
+pair<string, string>
+PCREKeyword::findExprInStr(const string &str, size_t start, size_t end)
+{
+    if (start>=end) throw KeywordError("The 'pcre' string is empty");
+
+    // There are two way to write the regular expression:
+    // Either between '/' charecters: "/regexp/"  (this is the default delimiter)
+    // Or use 'm' to set a delimiter: "mDregexpD" (here 'D' is used as the delimiter)
+    // The switch will set the parameter 'start' so the 'str[start]' is the first delimiter.
+    switch (str[start]) {
+        case '/': {
+            break;
+        }
+        case 'm': {
+            start++;
+            if (start>=end) {
+                throw KeywordError("Failed to detect a delimiter in the 'pcre' keyword regular expression");
+            }
+            break;
+        }
+        default:
+            KeywordError("Bad start for the 'pcre' regular expression");
+    }
+
+    size_t expr_end = str.find_last_of(str[start], end-1);
+    start++;
+    if (expr_end<=start) throw KeywordError("The 'pcre' regular expression is empty");
+
+    auto options_start = expr_end+1;
+    return make_pair(str.substr(start, expr_end-start), str.substr(options_start, end-options_start));
+}
+
+void
+PCREKeyword::parseOptions(const string &options)
+{
+    for (auto ch : options) {
+        switch (ch) {
+            case 'i': {
+                is_case_insensitive.setAttr("pcre", "nocase");
+                break;
+            }
+            case 'R': {
+                is_relative.setAttr("pcre", "relative");
+                break;
+            }
+            case 'm': {
+                is_multiline.setAttr("pcre", "multiline");
+                break;
+            }
+            case 's': {
+                is_dotall.setAttr("pcre", "dotall");
+                break;
+            }
+            case 'x': {
+                is_extended.setAttr("pcre", "extended");
+                break;
+            }
+            case 'E': {
+                is_dollar_endonly.setAttr("pcre", "dollar_endonly");
+                break;
+            }
+            case 'A': {
+                is_anchor.setAttr("pcre", "anchor");
+                break;
+            }
+            case 'G': {
+                is_ungreedy.setAttr("pcre", "ungreedy");
+                break;
+            }
+            default:
+                throw KeywordError("Unknown option " +  dumpHexChar(ch) + " in the 'pcre' keyword");
+        }
+    }
+}
+
+void
+PCREKeyword::compilePCRE(const string &expr)
+{
+    uint32_t options = PCRE2_NO_AUTO_CAPTURE;
+    if (is_case_insensitive) options |= PCRE2_CASELESS;
+    if (is_multiline)        options |= PCRE2_MULTILINE;
+    if (is_dotall)           options |= PCRE2_DOTALL;
+    if (is_extended)         options |= PCRE2_EXTENDED;
+    if (is_dollar_endonly)   options |= PCRE2_DOLLAR_ENDONLY;
+    if (is_anchor)           options |= PCRE2_ANCHORED;
+    if (is_ungreedy)         options |= PCRE2_UNGREEDY;
+
+    int error;
+    PCRE2_SIZE error_offset;
+    auto pattern = reinterpret_cast<PCRE2_SPTR>(expr.c_str());
+    pcre_machine.reset(pcre2_compile(pattern, expr.size(), options, &error, &error_offset, nullptr));
+    if (pcre_machine == nullptr) {
+        vector<u_char> msg;
+        msg.reserve(128);
+        pcre2_get_error_message(error, msg.data(), msg.capacity());
+
+        throw KeywordError(
+            "Failed to compile the 'pcre' at offset "
+            + to_string(error_offset)
+            + " with error: "
+            + reinterpret_cast<char *>(msg.data())
+        );
+    }
+
+    pcre_result.reset(pcre2_match_data_create_from_pattern(pcre_machine.get(), nullptr));
+    if (pcre_result == nullptr) {
+        throw KeywordError("Failed to allocate PCRE results container");
+    }
+
+    pcre_expr = expr;
+}
+
+static uint
+addOffset(uint offset, int add)
+{
+    if (add<0 && offset<static_cast<uint>(-add)) return 0;
+    return offset + add;
+}
+
+pair<uint, uint>
+PCREKeyword::getStartOffsetAndLength(uint buf_size, const I_KeywordRuntimeState *prev) const
+{
+    uint keyword_offset = is_relative?prev->getOffset(ctx):0;
+    uint start_offset = addOffset(keyword_offset, offset.evalAttr(prev));
+
+    if (start_offset>=buf_size) return make_pair(0, 0);
+
+    uint length = buf_size-start_offset;
+
+    if (depth.isSet()) {
+        length = min(length, static_cast<uint>(depth.evalAttr(prev)));
+    }
+
+    return make_pair(start_offset, length);
+}
+
+MatchStatus
+PCREKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    dbgAssert(pcre_machine!=nullptr) << "Trying to run on an uninitialized keyword 'pcre'";
+
+    auto part = Singleton::Consume<I_Environment>::by<KeywordComp>()->get<Buffer>(static_cast<string>(ctx));
+
+    if (!part.ok()) {
+        if (is_negative) {
+            return runNext(prev);
+        }
+        return MatchStatus::NoMatchFinal;
+    }
+
+    uint offset, length;
+    tie(offset, length) = getStartOffsetAndLength((*part).size(), prev);
+    auto buf = (*part).getPtr(offset, length);
+
+    if (!buf.ok()) {
+        dbgTrace(D_KEYWORD) << "Could not get the buffer for the 'pcre' keyword";
+        return MatchStatus::NoMatchFinal;
+    }
+    const unsigned char *ptr = *buf;
+
+    bool match_found = false;
+    uint buf_offset_found;
+    for (uint buf_pos = 0; buf_pos<length; buf_pos = buf_offset_found) {
+        dbgDebug(D_KEYWORD) << "Looking for expression: " << pcre_expr;
+        dbgTrace(D_KEYWORD) << "Running pcre_exec for expression: " << ptr;
+        int result = pcre2_match(
+            pcre_machine.get(),
+            ptr,
+            length,
+            buf_pos,
+            0,
+            pcre_result.get(),
+            nullptr
+        );
+
+        if (result<0) {
+            // No match (possiblely due to an error)
+            dbgDebug(D_KEYWORD) << "No match, possiblely due to an error in 'pcre_exec'";
+            break;
+        } else {
+            dbgDebug(D_KEYWORD) << "Match found";
+        }
+
+        if (is_negative) {
+            return isConstant()?MatchStatus::NoMatchFinal:MatchStatus::NoMatch;
+        }
+        match_found = true;
+        buf_offset_found = pcre2_get_ovector_pointer(pcre_result.get())[0];
+        OffsetRuntimeState new_offset(prev, ctx, offset+buf_offset_found);
+        auto next_keyword_result = runNext(&new_offset);
+        if (next_keyword_result!=MatchStatus::NoMatch) return next_keyword_result;
+        if (buf_offset_found<=buf_pos) buf_offset_found = buf_pos+1; // Deal with empty matches
+    }
+
+    // No matchs is a success for negative keywords
+    if (is_negative && !match_found) {
+        return runNext(prev);
+    }
+
+    // If there were no matchs and the keyword is an effected by other keywords, then we know that the rule won't match
+    if (isConstant() && !match_found) {
+        return MatchStatus::NoMatchFinal;
+    }
+
+    return MatchStatus::NoMatch;
+}
+
+unique_ptr<SingleKeyword>
+genPCREKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<PCREKeyword>(attr, known_vars);
+}

components/utils/keywords/sentinel_runtime_state.h

@@ -0,0 +1,13 @@
+#ifndef ___SENTINEL_RUNTIME_STATE_H__
+#define ___SENTINEL_RUNTIME_STATE_H__
+
+#include "single_keyword.h"
+
+class SentinelRuntimeState : public I_KeywordRuntimeState
+{
+public:
+    uint getOffset(const std::string &) const override;
+    uint getVariable(uint) const override;
+};
+
+#endif //  ___SENTINEL_RUNTIME_STATE_H__

components/utils/keywords/single_keyword.cc

@@ -0,0 +1,276 @@
+#include "single_keyword.h"
+
+#include <algorithm>
+
+using namespace std;
+
+void
+SingleKeyword::appendKeyword(unique_ptr<SingleKeyword> &&_next)
+{
+    if (next==nullptr) {
+        next = move(_next);
+    } else {
+        next->appendKeyword(move(_next));
+    }
+}
+
+MatchStatus
+SingleKeyword::runNext(const I_KeywordRuntimeState *curr) const
+{
+    if (next==nullptr) {
+        return MatchStatus::Match;
+    }
+    return next->isMatch(curr);
+}
+
+OffsetRuntimeState::OffsetRuntimeState(
+    const I_KeywordRuntimeState *_p,
+    const string &_ctx,
+    uint _offset)
+        :
+    prev(_p),
+    ctx(_ctx),
+    offset(_offset)
+{
+}
+
+uint
+OffsetRuntimeState::getOffset(const string &requested_ctx) const
+{
+    if (ctx==requested_ctx) return offset;
+    return prev->getOffset(requested_ctx);
+}
+
+uint
+OffsetRuntimeState::getVariable(uint requested_var_id) const
+{
+    return prev->getVariable(requested_var_id);
+}
+
+VariableRuntimeState::VariableRuntimeState(
+    const I_KeywordRuntimeState *_p,
+    uint _var_id,
+    uint _val)
+        :
+    prev(_p),
+    var_id(_var_id),
+    value(_val)
+{
+}
+
+uint
+VariableRuntimeState::getOffset(const string &requested_ctx) const
+{
+    return prev->getOffset(requested_ctx);
+}
+
+uint
+VariableRuntimeState::getVariable(uint requested_var_id) const
+{
+    if (var_id==requested_var_id) return value;
+    return prev->getVariable(requested_var_id);
+}
+
+uint
+VariablesMapping::addNewVariable(const string &param)
+{
+    auto iter = mapping.find(param);
+    if (iter==mapping.end()) {
+        mapping[param] = mapping.size();
+    }
+    return mapping[param];
+}
+
+Maybe<uint>
+VariablesMapping::getVariableId(const string &param) const
+{
+    auto iter = mapping.find(param);
+    if (iter==mapping.end()) {
+        return genError(string("Unknown parameter ")+param);
+    }
+    return iter->second;
+}
+
+void
+NumericAttr::setAttr(
+    const KeywordAttr &attr,
+    const VariablesMapping &known_vars,
+    const string &keyword_name,
+    const uint base,
+    bool is_unsigned_val)
+{
+    auto &vec = attr.getParams();
+    if (vec.size()!= 2) {
+        throw KeywordError("Malformed " + attr.getAttrName() + "' in the '" + keyword_name + "' keyword");
+    }
+
+    setAttr(attr.getAttrName(), vec[1], known_vars, keyword_name, base, is_unsigned_val);
+}
+
+void
+NumericAttr::setAttr(
+    const string &attr_name,
+    const string &param,
+    const VariablesMapping &known_vars,
+    const string &keyword_name,
+    const uint base,
+    bool is_unsigned_val)
+{
+    if (isSet()) {
+        throw KeywordError("Double definition of the '" + attr_name + "' in the '" + keyword_name + "' keyword");
+    }
+
+    if (is_unsigned_val && param[0]=='-') {
+        throw KeywordError(
+            "Negative constant '" +
+            param +
+            "' in the '" +
+            attr_name +
+            "' in the '" +
+            keyword_name +
+            "' keyword"
+        );
+    }
+
+    if (isdigit(param[0]) || param[0] == '-') {
+        status = Status::Const;
+        try {
+            size_t idx;
+            val = stol(param, &idx, base);
+            if (idx != param.length()) throw invalid_argument("");
+        }
+        catch (...) {
+            throw KeywordError(
+                "Malformed constant '" +
+                param +
+                "' in the '" +
+                attr_name +
+                "' in the '" +
+                keyword_name +
+                "' keyword"
+            );
+        }
+    } else {
+        status = Status::Var;
+        val = known_vars.getVariableId(param).unpack<KeywordError>(
+            "In " + keyword_name +
+            " in " + attr_name + ": "
+        );
+    }
+}
+
+int
+NumericAttr::evalAttr(const I_KeywordRuntimeState *prev) const
+{
+    if (status==Status::Var) {
+        return prev->getVariable(val);
+    }
+    return val;
+}
+
+void
+BoolAttr::setAttr(const KeywordAttr &attr, const string &keyword_name)
+{
+    if (attr.getParams().size()!=1) {
+        throw KeywordError("Malformed " + attr.getAttrName() + "' in the '" + keyword_name + "' keyword");
+    }
+
+    val = true;
+}
+
+void
+BoolAttr::setAttr(const string &keyword_name, const string &attr_name)
+{
+    if (val) throw KeywordError("Double definition of the '" + attr_name + "' in the '" + keyword_name + "' keyword");
+
+    val = true;
+}
+
+void
+CtxAttr::setAttr(const KeywordAttr &attr, const string &keyword_name)
+{
+    if (is_set) throw KeywordError("Double definition of the 'part' in the '" + keyword_name + "' keyword");
+    is_set = true;
+    auto vec = attr.getParams();
+    if (vec.size()!=2) throw KeywordError("Malformed 'part' in the '" + keyword_name + "' keyword");
+    ctx = vec[1];
+}
+
+const map<string, ComparisonAttr::CompId> ComparisonAttr::name_to_operator {
+    { "=",          CompId::EQUAL                 },
+    { "!=",         CompId::NOT_EQUAL             },
+    { "<",          CompId::LESS_THAN             },
+    { ">",          CompId::GREATER_THAN          },
+    { "<=",         CompId::LESS_THAN_OR_EQUAL    },
+    { ">=",         CompId::GREATER_THAN_OR_EQUAL }
+};
+
+Maybe<ComparisonAttr::CompId>
+ComparisonAttr::getComparisonByName(const string &name)
+{
+    auto iter = name_to_operator.find(name);
+    if (iter == name_to_operator.end()) {
+        return genError("Could not find the operator: " + name);
+    }
+    return iter->second;
+}
+
+void
+ComparisonAttr::setAttr(const string &param, const string &keyword_name)
+{
+    if (isSet()) {
+        throw KeywordError("Double definition of the comparison opearator in the '" + keyword_name + "' keyword");
+    }
+    is_set = true;
+    comp_val = getComparisonByName(param).unpack<KeywordError>(
+        "Unknown comparison operator in the '" + keyword_name  + "' keyword: "
+    );
+}
+
+bool
+ComparisonAttr::operator()(int first_val, int second_val) const
+{
+    switch (comp_val) {
+        case ComparisonAttr::CompId::EQUAL: {
+            return first_val == second_val;
+        }
+        case ComparisonAttr::CompId::NOT_EQUAL: {
+            return first_val != second_val;
+        }
+        case ComparisonAttr::CompId::LESS_THAN: {
+            return first_val < second_val;
+        }
+        case ComparisonAttr::CompId::GREATER_THAN: {
+            return first_val > second_val;
+        }
+        case ComparisonAttr::CompId::LESS_THAN_OR_EQUAL: {
+            return first_val <= second_val;
+        }
+        case ComparisonAttr::CompId::GREATER_THAN_OR_EQUAL: {
+            return first_val >= second_val;
+        }
+    }
+    dbgAssert(false) << "ComparisonAttr::operator found an invalid comparison operator";
+    return false;
+}
+
+using InitFunc = unique_ptr<SingleKeyword>(*)(const vector<KeywordAttr> &, VariablesMapping &);
+
+const map<string, InitFunc> initializers = {
+    {"data",          genDataKeyword        },
+    {"pcre",          genPCREKeyword        },
+    {"length",        genLengthKeyword      },
+    {"byte_extract",  genByteExtractKeyword },
+    {"compare",       genCompareKeyword     },
+    {"stateop",       genStateopKeyword     },
+    {"no_match",      genNoMatchKeyword     },
+    {"jump",          genJumpKeyword        }
+};
+
+unique_ptr<SingleKeyword>
+getKeywordByName(const KeywordParsed &keyword, VariablesMapping &known_vars)
+{
+    auto iter = initializers.find(keyword.getName());
+    if (iter==initializers.end()) throw KeywordError(keyword.getName() + " - unknown keyword type");
+    return iter->second(keyword.getAttr(), known_vars);
+}

components/utils/keywords/single_keyword.h

@@ -0,0 +1,300 @@
+#ifndef ___SINGLE_KEYWORD_H__
+#define ___SINGLE_KEYWORD_H__
+
+#include <map>
+#include <string>
+#include <vector>
+#include <memory>
+
+#include "buffer.h"
+#include "keyword_comp.h"
+#include "debug.h"
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+enum class MatchStatus { Match, NoMatch, NoMatchFinal };
+
+class KeywordError
+{
+public:
+    KeywordError(const std::string &str) : err(str)
+    {
+    }
+
+    const std::string &
+    getErr() const
+    {
+        return err;
+    }
+
+private:
+    std::string err;
+};
+
+class KeywordAttr
+{
+public:
+    KeywordAttr(const std::string &str);
+
+    const std::string&
+    getAttrName() const
+    {
+        return params[0];
+    }
+
+    const std::vector<std::string> &
+    getParams() const
+    {
+        return params;
+    }
+
+private:
+    std::vector<std::string> params;
+};
+
+class KeywordParsed
+{
+public:
+    KeywordParsed(const std::string &keyword);
+
+    const std::string &
+    getName() const
+    {
+        return name;
+    }
+
+    const std::vector<KeywordAttr> &
+    getAttr() const
+    {
+        return attr;
+    }
+
+private:
+    std::string name;
+    std::vector<KeywordAttr> attr;
+};
+
+class I_KeywordRuntimeState
+{
+public:
+    virtual uint getOffset(const std::string &ctx) const = 0;
+    virtual uint getVariable(uint requested_var_id) const = 0;
+protected:
+    virtual ~I_KeywordRuntimeState()
+    {
+    }
+};
+
+class OffsetRuntimeState : public I_KeywordRuntimeState
+{
+public:
+    OffsetRuntimeState(const I_KeywordRuntimeState *prev, const std::string &ctx, uint offset);
+    virtual ~OffsetRuntimeState()
+    {
+    }
+    uint getOffset(const std::string &requested_ctx_id) const override;
+    uint getVariable(uint requested_var_id) const override;
+
+private:
+    const I_KeywordRuntimeState *prev;
+    std::string ctx;
+    uint offset;
+};
+
+class VariableRuntimeState : public I_KeywordRuntimeState
+{
+public:
+    VariableRuntimeState(const I_KeywordRuntimeState *prev, uint var_id, uint val);
+    virtual ~VariableRuntimeState()
+    {
+    }
+
+    uint getOffset(const std::string &requested_ctx_id) const override;
+    uint getVariable(uint requested_var_id) const override;
+
+private:
+    const I_KeywordRuntimeState *prev;
+    uint var_id;
+    uint value;
+};
+
+class VariablesMapping
+{
+public:
+    uint addNewVariable(const std::string &name);
+    Maybe<uint> getVariableId(const std::string &name) const;
+
+private:
+    std::map<std::string, uint> mapping;
+};
+
+class NumericAttr
+{
+    enum class Status { Unset, Const, Var };
+public:
+    void setAttr(
+        const KeywordAttr &attr,
+        const VariablesMapping &known_vars,
+        const std::string &keyword_name,
+        const uint base = 10,
+        bool is_unsigned_val = false);
+
+    void setAttr(
+        const std::string &attr_name,
+        const std::string &param,
+        const VariablesMapping &known_vars,
+        const std::string &keyword_name,
+        const uint base = 10,
+        bool is_unsigned_val = false);
+
+    int evalAttr(const I_KeywordRuntimeState *prev) const;
+
+    bool
+    isConstant() const
+    {
+        return status!=Status::Var;
+    }
+
+    bool
+    isSet() const
+    {
+        return status!=Status::Unset;
+    }
+
+private:
+    Status status = Status::Unset;
+    int val = 0;
+};
+
+class BoolAttr
+{
+public:
+    void setAttr(const KeywordAttr &attr, const std::string &keyword_name);
+    void setAttr(const std::string &keyword_name, const std::string &attr_name);
+
+    operator bool() const
+    {
+        return val;
+    }
+
+private:
+    bool val = false;
+};
+
+class CtxAttr
+{
+public:
+    void setAttr(const KeywordAttr &attr, const std::string &keyword_name);
+
+    operator std::string() const
+    {
+        if (!is_set) {
+            auto env = Singleton::Consume<I_Environment>::by<KeywordComp>();
+            auto default_ctx = env->get<std::string>(I_KeywordsRule::getKeywordsRuleTag());
+            if (default_ctx.ok()) return *default_ctx;
+            dbgError(D_KEYWORD) << "Running keyword rule without specific context and without default";
+            return "Missing Default Context";
+        }
+        return ctx;
+    }
+
+private:
+    std::string ctx;
+    bool is_set = false;
+};
+
+class ComparisonAttr
+{
+public:
+    enum class CompId
+    {
+        EQUAL,
+        NOT_EQUAL,
+        LESS_THAN,
+        GREATER_THAN,
+        LESS_THAN_OR_EQUAL,
+        GREATER_THAN_OR_EQUAL
+    };
+
+    void setAttr(const std::string &param, const std::string &keyword_name);
+    bool operator()(int first_val, int second_val) const;
+
+    bool
+    isSet() const
+    {
+        return is_set;
+    }
+
+private:
+    Maybe<CompId> getComparisonByName(const std::string &name);
+
+    const static std::map<std::string, CompId> name_to_operator;
+    bool is_set = false;
+    CompId comp_val;
+};
+
+class SingleKeyword
+{
+public:
+    SingleKeyword()
+    {
+    }
+
+    virtual ~SingleKeyword()
+    {
+    }
+
+    MatchStatus runNext(const I_KeywordRuntimeState *curr) const;
+    void appendKeyword(std::unique_ptr<SingleKeyword> &&_next);
+    virtual MatchStatus isMatch(const I_KeywordRuntimeState *prev) const = 0;
+
+private:
+    std::unique_ptr<SingleKeyword> next;
+};
+
+std::unique_ptr<SingleKeyword> genDataKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genPCREKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genLengthKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genByteExtractKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genCompareKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genStateopKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genNoMatchKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> genJumpKeyword(
+    const std::vector<KeywordAttr> &attr,
+    VariablesMapping &vars
+);
+
+std::unique_ptr<SingleKeyword> getKeywordByName(
+    const KeywordParsed &parsed_data,
+    VariablesMapping &vars
+);
+
+#endif // ___SINGLE_KEYWORD_H__

components/utils/keywords/stateop_keyword.cc

@@ -0,0 +1,145 @@
+#include "single_keyword.h"
+#include "table_opaque.h"
+#include "debug.h"
+#include "flags.h"
+
+#include <map>
+#include <strings.h>
+
+#include "cereal/types/set.hpp"
+
+using namespace std;
+
+USE_DEBUG_FLAG(D_KEYWORD);
+
+class StateopKeyword : public SingleKeyword
+{
+public:
+    explicit StateopKeyword(const vector<KeywordAttr> &attr, VariablesMapping &vars);
+    MatchStatus isMatch(const I_KeywordRuntimeState* prev) const override;
+
+private:
+    enum class Operation { ISSET, SET, UNSET, COUNT };
+    using OpFlags = Flags<Operation>;
+
+    void
+    setState(const KeywordAttr &attr, const VariablesMapping &)
+    {
+        auto &var_name_param = attr.getParams();
+
+        if (var_name_param.size() != 2) {
+            throw KeywordError("More than one element in the state name in the 'stateop' keyword");
+        }
+
+        var_name = var_name_param[1];
+    }
+
+    void
+    setTesting(const KeywordAttr &, const VariablesMapping &)
+    {
+        if (!mode.empty()) throw KeywordError("Redefining 'stateop' keyword operation");
+        mode.setFlag(Operation::ISSET);
+    }
+
+    void
+    setSetting(const KeywordAttr &, const VariablesMapping &)
+    {
+        if (!mode.empty()) throw KeywordError("Redefining 'stateop' keyword operation");
+        mode.setFlag(Operation::SET);
+    }
+
+    void
+    setUnsetting(const KeywordAttr &, const VariablesMapping &)
+    {
+        if (!mode.empty()) throw KeywordError("Redefining 'stateop' keyword operation");
+        mode.setFlag(Operation::UNSET);
+    }
+
+    string  var_name;
+    OpFlags mode;
+
+    static const map<string, void(StateopKeyword::*)(const KeywordAttr &, const VariablesMapping &)> setops;
+};
+
+const map<string, void(StateopKeyword::*)(const KeywordAttr &, const VariablesMapping &)> StateopKeyword::setops = {
+    { "isset", &StateopKeyword::setTesting   },
+    { "set",   &StateopKeyword::setSetting   },
+    { "unset", &StateopKeyword::setUnsetting },
+    { "state", &StateopKeyword::setState     }
+};
+
+StateopKeyword::StateopKeyword(const vector<KeywordAttr> &attrs, VariablesMapping &vars)
+{
+    if (attrs.size() != 2) throw KeywordError("Invalid number of attributes in the 'stateop' keyword");
+    
+    for (uint i = 0; i < attrs.size(); i++) {
+        auto curr = setops.find(attrs[i].getAttrName());
+        if (curr == setops.end()) {
+            throw KeywordError("Unknown attribute '" + attrs[i].getAttrName() + "' in the 'stateop' keyword");
+        }
+        auto set_func = curr->second;
+        (this->*set_func)(attrs[i], vars);
+    }
+
+    if (var_name == "" || mode.empty()) {
+        throw KeywordError("Bad 'stateop' attribute configuration");
+    }
+}
+
+class KeywordStateop : public TableOpaqueSerialize<KeywordStateop>
+{
+public:
+    KeywordStateop() : TableOpaqueSerialize<KeywordStateop>(this) {}
+
+    bool hasVariable(const string &state) { return states.find(state) != states.end(); }
+    void addVariable(const string &state) { states.insert(state); }
+    void removeVariable(const string &state) { states.erase(state); }
+
+    // LCOV_EXCL_START - sync functions, can only be tested once the sync module exists
+    template <typename T>
+    void
+    serialize(T &ar, uint32_t)
+    {
+        ar(states);
+    }
+
+    static std::string name() { return "KeywordStateop"; }
+    static std::unique_ptr<TableOpaqueBase> prototype() { return std::make_unique<KeywordStateop>(); }
+    static uint currVer() { return 0; }
+    static uint minVer() { return 0; }
+    // LCOV_EXCL_STOP
+
+private:
+    set<string> states;
+};
+
+MatchStatus
+StateopKeyword::isMatch(const I_KeywordRuntimeState *prev) const
+{
+    auto table = Singleton::Consume<I_Table>::by<KeywordComp>();
+
+    if (mode.isSet(Operation::ISSET)) {
+        if (!table->hasState<KeywordStateop>()) return MatchStatus::NoMatchFinal;
+        auto &state = table->getState<KeywordStateop>();
+        if (state.hasVariable(var_name)) return runNext(prev);
+        else return MatchStatus::NoMatchFinal;
+    } else if (mode.isSet(Operation::SET)) {
+        if (!table->hasState<KeywordStateop>()) table->createState<KeywordStateop>();
+        table->getState<KeywordStateop>().addVariable(var_name);
+        return runNext(prev);
+    } else if (mode.isSet(Operation::UNSET)) {
+        if (table->hasState<KeywordStateop>()) table->getState<KeywordStateop>().removeVariable(var_name);
+        return runNext(prev);
+    } else {
+        dbgAssert(false) << "Impossible 'stateop' keyword without operation";
+    }
+
+    // If there was no matches and the keyword is effected by other keywords, then we know that the rule won't match
+    return MatchStatus::NoMatchFinal;
+}
+
+unique_ptr<SingleKeyword>
+genStateopKeyword(const vector<KeywordAttr> &attr, VariablesMapping &known_vars)
+{
+    return make_unique<StateopKeyword>(attr, known_vars);
+}