aboutsummaryrefslogtreecommitdiff
path: root/src/openvic-dataloader/csv/CsvGrammar.hpp
diff options
context:
space:
mode:
author Spartan322 <Megacake1234@gmail.com>2023-07-28 06:52:00 +0200
committer Spartan322 <Megacake1234@gmail.com>2023-09-02 14:28:21 +0200
commit7440a5d1433eec4bf87e3723022db187e7f61b1a (patch)
tree2bb062c320fa2227b18956617b94d0e8800420d8 /src/openvic-dataloader/csv/CsvGrammar.hpp
parente941573f47fb867ff75c8a2cf78302b754ffbeee (diff)
Rework Grammar and Parser
Add proper headless binary construction: Includes basic validation Add Error and Warning structs to Parser Add FileNode pointer getter to Parser Change all `char8_t*` and `const char8_t` to `const char*` in Parser Add Parser move operators and Parser deconstructor Add BufferHandler PIMPL object to Parser Add UTF-8 file Warning to v2script Add proper Grammar value retrieval Add AbstractSyntaxTree for v2script data parser: Has compile-time embedded type information accessible at compile-time and runtime Has Tab-based print functionality Fix wrong environment reference for headless construction in SConstruct Add error retrieval Add BasicCallbackOStreamBuffer for callback streaming Add CallbackStreamBuffer for char Add CallbackWStreamBuffer for wchar_t Add BasicCallbackStream Add CallbackStream for char Add CallbackWStream for wchar_t Add grammar for events and decisions Add event_parse to Parser Add decision_parse to Parser Add .clang-format Ignore dirty lexy module Add CSV parser and grammar: Creates std::vector<csv::LineObject> for a list of lines Add BasicParser and BasicBufferHandler to reduce code reduplication
Diffstat (limited to 'src/openvic-dataloader/csv/CsvGrammar.hpp')
-rw-r--r--src/openvic-dataloader/csv/CsvGrammar.hpp129
1 files changed, 129 insertions, 0 deletions
diff --git a/src/openvic-dataloader/csv/CsvGrammar.hpp b/src/openvic-dataloader/csv/CsvGrammar.hpp
new file mode 100644
index 0000000..edce97b
--- /dev/null
+++ b/src/openvic-dataloader/csv/CsvGrammar.hpp
@@ -0,0 +1,129 @@
+#pragma once
+
+#include <initializer_list>
+#include <string>
+#include <tuple>
+#include <type_traits>
+#include <vector>
+
+#include <openvic-dataloader/csv/LineObject.hpp>
+
+#include <lexy/callback.hpp>
+#include <lexy/dsl.hpp>
+
+// Grammar Definitions //
+namespace ovdl::csv::grammar {
+ struct StringValue {
+ static constexpr auto escaped_symbols = lexy::symbol_table<char> //
+ .map<'"'>('"')
+ .map<'\''>('\'')
+ .map<'\\'>('\\')
+ .map<'/'>('/')
+ .map<'b'>('\b')
+ .map<'f'>('\f')
+ .map<'n'>('\n')
+ .map<'r'>('\r')
+ .map<'t'>('\t');
+ /// This doesn't actually do anything, so this might to be manually parsed if vic2's CSV parser creates a " from ""
+ static constexpr auto escaped_quote = lexy::symbol_table<char> //
+ .map<'"'>('"');
+ static constexpr auto rule = [] {
+ // Arbitrary code points
+ auto c = -lexy::dsl::lit_c<'"'>;
+
+ auto back_escape = lexy::dsl::backslash_escape //
+ .symbol<escaped_symbols>();
+
+ auto quote_escape = lexy::dsl::escape(lexy::dsl::lit_c<'"'>) //
+ .symbol<escaped_quote>();
+
+ return lexy::dsl::quoted(c, back_escape, quote_escape);
+ }();
+
+ static constexpr auto value = lexy::as_string<std::string>;
+ };
+
+ template<auto Sep>
+ struct PlainValue {
+ static constexpr auto rule = lexy::dsl::identifier(-(Sep / lexy::dsl::lit_c<'\n'>));
+ static constexpr auto value = lexy::as_string<std::string>;
+ };
+
+ template<auto Sep>
+ struct Value {
+ static constexpr auto rule = lexy::dsl::p<StringValue> | lexy::dsl::p<PlainValue<Sep>>;
+ static constexpr auto value = lexy::forward<std::string>;
+ };
+
+ template<auto Sep>
+ struct SeperatorCount {
+ static constexpr auto rule = lexy::dsl::list(Sep);
+ static constexpr auto value = lexy::count;
+ };
+
+ template<auto Sep>
+ struct LineEnd {
+ static constexpr auto rule = lexy::dsl::list(lexy::dsl::p<Value<Sep>>, lexy::dsl::trailing_sep(lexy::dsl::p<SeperatorCount<Sep>>));
+ static constexpr auto value = lexy::fold_inplace<csv::LineObject>(
+ std::initializer_list<csv::LineObject::value_type> {},
+ [](csv::LineObject& result, auto&& arg) {
+ if constexpr (std::is_same_v<std::decay_t<decltype(arg)>, std::size_t>) {
+ // Count seperators, adds to previous value, making it a position
+ using position_type = csv::LineObject::position_type;
+ result.emplace_back(static_cast<position_type>(arg + std::get<0>(result.back())), "");
+ } else {
+ if (result.empty()) result.emplace_back(0u, LEXY_MOV(arg));
+ else {
+ auto& [pos, value] = result.back();
+ value = arg;
+ }
+ }
+ });
+ };
+
+ template<auto Sep>
+ struct Line {
+
+ static constexpr auto suffix_setter(csv::LineObject& line) {
+ auto& [position, value] = line.back();
+ if (value.empty()) {
+ line.set_suffix_end(position);
+ line.pop_back();
+ } else {
+ line.set_suffix_end(position + 1);
+ }
+ };
+
+ static constexpr auto rule = lexy::dsl::p<LineEnd<Sep>> | lexy::dsl::p<SeperatorCount<Sep>> >> lexy::dsl::p<LineEnd<Sep>>;
+ static constexpr auto value =
+ lexy::callback<csv::LineObject>(
+ [](csv::LineObject&& line) {
+ suffix_setter(line);
+ return LEXY_MOV(line);
+ },
+ [](std::size_t prefix_count, csv::LineObject&& line) {
+ line.set_prefix_end(prefix_count);
+ // position needs to be adjusted to prefix
+ for (auto& [position, value] : line) {
+ position += prefix_count;
+ }
+ suffix_setter(line);
+ return LEXY_MOV(line);
+ });
+ };
+
+ template<auto Sep>
+ struct File {
+ static constexpr auto rule =
+ lexy::dsl::whitespace(lexy::dsl::newline) +
+ lexy::dsl::opt(lexy::dsl::list(lexy::dsl::p<Line<Sep>>, lexy::dsl::trailing_sep(lexy::dsl::eol)));
+
+ static constexpr auto value = lexy::as_list<std::vector<csv::LineObject>>;
+ };
+
+ using CommaFile = File<lexy::dsl::lit_c<','>>;
+ using ColonFile = File<lexy::dsl::lit_c<':'>>;
+ using SemiColonFile = File<lexy::dsl::lit_c<';'>>;
+ using TabFile = File<lexy::dsl::lit_c<'\t'>>;
+ using BarFile = File<lexy::dsl::lit_c<'|'>>;
+} \ No newline at end of file