# HG changeset patch # User František Kučera # Date 1606002240 -3600 # Node ID f031a4dc7c52bd3439e99a8f118b967d53435006 # Parent 3876a9c56a6601bf8a1a8521d882b0c92a86d627 add options: --enable-sections --enable-subkeys --enable-comments --enable-line-numbers --enable-event-numbers diff -r 3876a9c56a66 -r f031a4dc7c52 bash-completion.sh --- a/bash-completion.sh Sat Nov 21 20:09:18 2020 +0100 +++ b/bash-completion.sh Sun Nov 22 00:44:00 2020 +0100 @@ -22,10 +22,25 @@ w2=${COMP_WORDS[COMP_CWORD-2]} w3=${COMP_WORDS[COMP_CWORD-3]} + BOOLEAN_VALUES=( + "true" + "false" + ) + if [[ "$w1" == "--relation" && "x$w0" == "x" ]]; then COMPREPLY=("''") + elif [[ "$w1" == "--enable-sections" ]]; then COMPREPLY=($(compgen -W "${BOOLEAN_VALUES[*]}" -- "$w0")) + elif [[ "$w1" == "--enable-subkeys" ]]; then COMPREPLY=($(compgen -W "${BOOLEAN_VALUES[*]}" -- "$w0")) + elif [[ "$w1" == "--enable-comments" ]]; then COMPREPLY=($(compgen -W "${BOOLEAN_VALUES[*]}" -- "$w0")) + elif [[ "$w1" == "--enable-line-numbers" ]]; then COMPREPLY=($(compgen -W "${BOOLEAN_VALUES[*]}" -- "$w0")) + elif [[ "$w1" == "--enable-event-numbers" ]]; then COMPREPLY=($(compgen -W "${BOOLEAN_VALUES[*]}" -- "$w0")) else OPTIONS=( "--relation" + "--enable-sections" + "--enable-subkeys" + "--enable-comments" + "--enable-line-numbers" + "--enable-event-numbers" ) COMPREPLY=($(compgen -W "${OPTIONS[*]}" -- "$w0")) fi diff -r 3876a9c56a66 -r f031a4dc7c52 src/CLIParser.h --- a/src/CLIParser.h Sat Nov 21 20:09:18 2020 +0100 +++ b/src/CLIParser.h Sun Nov 22 00:44:00 2020 +0100 @@ -37,9 +37,23 @@ else throw relpipe::cli::RelpipeCLIException(L"Missing CLI argument" + (i > 0 ? (L" after " + arguments[i - 1]) : L""), relpipe::cli::CLI::EXIT_CODE_BAD_CLI_ARGUMENTS); } + /** + * TODO: use a common method + */ + bool parseBoolean(const relpipe::writer::string_t& value) { + if (value == L"true") return true; + else if (value == L"false") return false; + else throw relpipe::cli::RelpipeCLIException(L"Unable to parse boolean value: " + value + L" (expecting true or false)", relpipe::cli::CLI::EXIT_CODE_BAD_CLI_ARGUMENTS); + } + public: static const relpipe::writer::string_t OPTION_RELATION; + static const relpipe::writer::string_t OPTION_ENABLE_SECTIONS; + static const relpipe::writer::string_t OPTION_ENABLE_SUBKEYS; + static const relpipe::writer::string_t OPTION_ENABLE_COMMENTS; + static const relpipe::writer::string_t OPTION_ENABLE_LINE_NUMBERS; + static const relpipe::writer::string_t OPTION_ENABLE_EVENT_NUMBERS; Configuration parse(const std::vector& arguments) { Configuration c; @@ -47,9 +61,13 @@ for (int i = 0; i < arguments.size();) { relpipe::writer::string_t option = readNext(arguments, i); - if (option == OPTION_RELATION) { - c.relation = readNext(arguments, i); - } else throw relpipe::cli::RelpipeCLIException(L"Unsupported CLI option: " + option, relpipe::cli::CLI::EXIT_CODE_BAD_CLI_ARGUMENTS); + if (option == OPTION_RELATION) c.relation = readNext(arguments, i); + else if (option == OPTION_ENABLE_SECTIONS) c.enableSections = parseBoolean(readNext(arguments, i)); + else if (option == OPTION_ENABLE_SUBKEYS) c.enableSubkeys = parseBoolean(readNext(arguments, i)); + else if (option == OPTION_ENABLE_COMMENTS) c.enableComments = parseBoolean(readNext(arguments, i)); + else if (option == OPTION_ENABLE_LINE_NUMBERS) c.enableLineNumbers = parseBoolean(readNext(arguments, i)); + else if (option == OPTION_ENABLE_EVENT_NUMBERS) c.enableEventNumbers = parseBoolean(readNext(arguments, i)); + else throw relpipe::cli::RelpipeCLIException(L"Unsupported CLI option: " + option, relpipe::cli::CLI::EXIT_CODE_BAD_CLI_ARGUMENTS); } return c; @@ -60,6 +78,11 @@ }; const relpipe::writer::string_t CLIParser::OPTION_RELATION = L"--relation"; +const relpipe::writer::string_t CLIParser::OPTION_ENABLE_SECTIONS = L"--enable-sections"; +const relpipe::writer::string_t CLIParser::OPTION_ENABLE_SUBKEYS = L"--enable-subkeys"; +const relpipe::writer::string_t CLIParser::OPTION_ENABLE_COMMENTS = L"--enable-comments"; +const relpipe::writer::string_t CLIParser::OPTION_ENABLE_LINE_NUMBERS = L"--enable-line-numbers"; +const relpipe::writer::string_t CLIParser::OPTION_ENABLE_EVENT_NUMBERS = L"--enable-event-numbers"; } } diff -r 3876a9c56a66 -r f031a4dc7c52 src/Configuration.h --- a/src/Configuration.h Sat Nov 21 20:09:18 2020 +0100 +++ b/src/Configuration.h Sun Nov 22 00:44:00 2020 +0100 @@ -29,6 +29,11 @@ class Configuration { public: relpipe::writer::string_t relation = L"ini"; + relpipe::writer::boolean_t enableLineNumbers = false; + relpipe::writer::boolean_t enableEventNumbers = false; + relpipe::writer::boolean_t enableSections = true; + relpipe::writer::boolean_t enableSubkeys = false; + relpipe::writer::boolean_t enableComments = false; virtual ~Configuration() { } diff -r 3876a9c56a66 -r f031a4dc7c52 src/INICommand.cpp --- a/src/INICommand.cpp Sat Nov 21 20:09:18 2020 +0100 +++ b/src/INICommand.cpp Sun Nov 22 00:44:00 2020 +0100 @@ -66,10 +66,13 @@ void startDocument() override { vector metadata; - metadata.push_back({L"section", TypeId::STRING}); + if (configuration.enableLineNumbers) metadata.push_back({L"line", TypeId::INTEGER}); + if (configuration.enableEventNumbers) metadata.push_back({L"event", TypeId::INTEGER}); + if (configuration.enableSections) metadata.push_back({L"section", TypeId::STRING}); metadata.push_back({L"key", TypeId::STRING}); - metadata.push_back({L"subkey", TypeId::STRING}); + if (configuration.enableSubkeys) metadata.push_back({L"subkey", TypeId::STRING}); metadata.push_back({L"value", TypeId::STRING}); + if (configuration.enableComments) metadata.push_back({L"comment", TypeId::STRING}); writer->startRelation(configuration.relation, metadata, true); }; @@ -77,23 +80,31 @@ currentSection.clear(); }; - void startSection(const std::string& name) override { - currentSection.push_back(name); + void startSection(const SectionStartEvent& event) override { + currentSection.push_back(event.name); }; void endSection() override { currentSection.pop_back(); }; - void entry(const std::string& key, const std::string& subkey, const std::string& value) override { - writer->writeAttribute(convertor.from_bytes(getCurrentSectionFullName())); + void entry(const EntryEvent& event) override { + if (configuration.enableLineNumbers) writer->writeAttribute(&event.lineNumber, typeid (event.lineNumber)); + if (configuration.enableEventNumbers) writer->writeAttribute(&event.eventNumber, typeid (event.eventNumber)); + + std::string section = getCurrentSectionFullName(); + std::string key = configuration.enableSubkeys ? event.key : event.fullKey; + + if (configuration.enableSections) writer->writeAttribute(convertor.from_bytes(section)); + else if (section.size()) key = section + "/" + key; writer->writeAttribute(convertor.from_bytes(key)); - writer->writeAttribute(convertor.from_bytes(subkey)); - writer->writeAttribute(convertor.from_bytes(value)); + if (configuration.enableSubkeys) writer->writeAttribute(convertor.from_bytes(event.subKey)); + + writer->writeAttribute(convertor.from_bytes(event.value)); + if (configuration.enableComments) writer->writeAttribute(convertor.from_bytes(event.comment)); }; // TODO: handle also comments and whitespace (to allow lossless transformation from INI and back to INI) - // TODO: make subkeys (in [] brackets in the key) optional/configurable }; diff -r 3876a9c56a66 -r f031a4dc7c52 src/lib/INIContentHandler.h --- a/src/lib/INIContentHandler.h Sat Nov 21 20:09:18 2020 +0100 +++ b/src/lib/INIContentHandler.h Sun Nov 22 00:44:00 2020 +0100 @@ -20,10 +20,31 @@ class INIContentHandler { public: + + class Event { + public: + int64_t eventNumber = -1; + int64_t lineNumber = -1; + std::string comment; + }; + + class SectionStartEvent : public Event { + public: + std::string name; + }; + + class EntryEvent : public Event { + public: + std::string key; + std::string subKey; + std::string fullKey; + std::string value; + }; + virtual ~INIContentHandler() = default; virtual void startDocument() = 0; virtual void endDocument() = 0; - virtual void startSection(const std::string& name) = 0; + virtual void startSection(const SectionStartEvent& event) = 0; virtual void endSection() = 0; - virtual void entry(const std::string& key, const std::string& subkey, const std::string& value) = 0; + virtual void entry(const EntryEvent& event) = 0; }; \ No newline at end of file diff -r 3876a9c56a66 -r f031a4dc7c52 src/lib/INIReader.cpp --- a/src/lib/INIReader.cpp Sat Nov 21 20:09:18 2020 +0100 +++ b/src/lib/INIReader.cpp Sun Nov 22 00:44:00 2020 +0100 @@ -40,30 +40,43 @@ std::regex whitespacePattrern("\\s*"); std::regex commentPattrern("\\s*(;|#)\\s*(.*)"); std::regex sectionPattrern("\\s*\\[\\s*([^\\]]+)\\s*\\]\\s*"); - std::regex entryQuotesPattrern("\\s*([^=\\]]+?[^=\\s\\]]*)(\\[([^\\]]+)\\])?\\s*=\\s*\"([^']+)\"\\s*((;|#)\\s*(.*))?"); - std::regex entryApostrophesPattrern("\\s*([^=\\]]+?[^=\\s\\]]*)(\\[([^\\]]+)\\])?\\s*=\\s*'([^']+)'\\s*((;|#)\\s*(.*))?"); - std::regex entryPlainPattrern("\\s*([^=\\]]+?[^=\\s\\]]*)(\\[([^\\]]+)\\])?\\s*=\\s*(.*)"); + std::regex entryQuotesPattrern(/***/"\\s*(([^=\\]]+?[^=\\s\\]]*)(\\[([^\\]]+)\\])?)\\s*=\\s*\"([^']+)\"\\s*((;|#)\\s*(.*))?"); + std::regex entryApostrophesPattrern("\\s*(([^=\\]]+?[^=\\s\\]]*)(\\[([^\\]]+)\\])?)\\s*=\\s*'([^']+)'\\s*((;|#)\\s*(.*))?"); + std::regex entryPlainPattrern("\\s*(([^=\\]]+?[^=\\s\\]]*)(\\[([^\\]]+)\\])?)\\s*=\\s*(.*)"); std::smatch match; - std::string section; + bool inSection = false; std::string line; + int lineNumber = 0; + int eventNumber = 0; + while (std::getline(input, line)) { + lineNumber++; if (std::regex_match(line, match, whitespacePattrern)) { // TODO: support also whitespace } else if (std::regex_match(line, match, commentPattrern)) { // TODO: support also comments + emit also the comment style (;/#) } else if (std::regex_match(line, match, sectionPattrern)) { - if (section.size()) for (INIContentHandler* handler : handlers) handler->endSection(); - section = match[1]; - for (INIContentHandler* handler : handlers) handler->startSection(section); - } else if (std::regex_match(line, match, entryQuotesPattrern) || std::regex_match(line, match, entryApostrophesPattrern)) { + if (inSection) for (INIContentHandler* handler : handlers) handler->endSection(); + INIContentHandler::SectionStartEvent event; + event.lineNumber = lineNumber; + event.eventNumber = ++eventNumber; + event.name = match[1]; // TODO: support also comments + emit also the comment style (;/#) + for (INIContentHandler* handler : handlers) handler->startSection(event); + } else if (std::regex_match(line, match, entryQuotesPattrern) || std::regex_match(line, match, entryApostrophesPattrern) || std::regex_match(line, match, entryPlainPattrern)) { + INIContentHandler::EntryEvent event; + event.lineNumber = lineNumber; + event.eventNumber = ++eventNumber; + event.key = match[2]; + event.subKey = match[4]; + event.fullKey = match[1]; + event.value = match[5]; + if (match.size() == 9) event.comment = match[8]; // TODO: emit also the quote style ('/"/) and surrounding whitespace - for (INIContentHandler* handler : handlers) handler->entry(match[1], match[3], match[4]); - } else if (std::regex_match(line, match, entryPlainPattrern)) { - for (INIContentHandler* handler : handlers) handler->entry(match[1], match[3], match[4]); + for (INIContentHandler* handler : handlers) handler->entry(event); } else { // TODO: warning, error, or support unknown content } @@ -77,10 +90,9 @@ // TODO: support also nested sections – hierarchy // TODO: support also option for alternative key-value separator (: instead of =) // TODO: support also other encodings (currently only UTF-8 is supported) - // TODO: emit line numbers and/or event order? } - if (section.size()) for (INIContentHandler* handler : handlers) handler->endSection(); + if (inSection) for (INIContentHandler* handler : handlers) handler->endSection(); for (INIContentHandler* handler : handlers) handler->endDocument(); } diff -r 3876a9c56a66 -r f031a4dc7c52 src/lib/INIReader.h --- a/src/lib/INIReader.h Sat Nov 21 20:09:18 2020 +0100 +++ b/src/lib/INIReader.h Sun Nov 22 00:44:00 2020 +0100 @@ -30,4 +30,4 @@ virtual void addHandler(INIContentHandler* handler) = 0; virtual void process() = 0; static INIReader* create(std::istream& input); -}; \ No newline at end of file +};