From 67b63f91f2668ec65e1791ecfeb24a53c389f1ec Mon Sep 17 00:00:00 2001 From: Karen Arutyunov Date: Thu, 8 Dec 2016 18:51:22 +0300 Subject: Add line_regex --- unit-tests/test/script/buildfile | 2 +- unit-tests/test/script/parser/buildfile | 2 +- unit-tests/test/script/regex/buildfile | 12 ++ unit-tests/test/script/regex/driver.cxx | 252 ++++++++++++++++++++++++++++++++ 4 files changed, 266 insertions(+), 2 deletions(-) create mode 100644 unit-tests/test/script/regex/buildfile create mode 100644 unit-tests/test/script/regex/driver.cxx (limited to 'unit-tests') diff --git a/unit-tests/test/script/buildfile b/unit-tests/test/script/buildfile index 6b303e1..c5306b1 100644 --- a/unit-tests/test/script/buildfile +++ b/unit-tests/test/script/buildfile @@ -2,6 +2,6 @@ # copyright : Copyright (c) 2014-2016 Code Synthesis Ltd # license : MIT; see accompanying LICENSE file -d = lexer/ parser/ +d = lexer/ parser/ regex/ ./: $d include $d diff --git a/unit-tests/test/script/parser/buildfile b/unit-tests/test/script/parser/buildfile index ad32494..db75a4f 100644 --- a/unit-tests/test/script/parser/buildfile +++ b/unit-tests/test/script/parser/buildfile @@ -11,7 +11,7 @@ src = token lexer parser diagnostics utility variable name context target \ scope prerequisite file module operation rule b-options algorithm search \ filesystem function functions-builtin functions-path functions-process-path \ functions-string config/{utility init operation} dump types-parsers \ -test/{target script/{token lexer parser script}} \ +test/{target script/{token lexer parser regex script}} \ scheduler exe{driver}: cxx{driver} ../../../../build2/cxx{$src} $libs \ diff --git a/unit-tests/test/script/regex/buildfile b/unit-tests/test/script/regex/buildfile new file mode 100644 index 0000000..26c759a --- /dev/null +++ b/unit-tests/test/script/regex/buildfile @@ -0,0 +1,12 @@ +# file : unit-tests/test/script/regex/buildfile +# copyright : Copyright (c) 2014-2016 Code Synthesis Ltd +# license : MIT; see accompanying LICENSE file + +#@@ Temporary until we get utility library support. +# +import libs = libbutl%lib{butl} +src = test/script/regex + +exe{driver}: cxx{driver} ../../../../build2/cxx{$src} $libs + +include ../../../../build2/ diff --git a/unit-tests/test/script/regex/driver.cxx b/unit-tests/test/script/regex/driver.cxx new file mode 100644 index 0000000..ca09048 --- /dev/null +++ b/unit-tests/test/script/regex/driver.cxx @@ -0,0 +1,252 @@ +// file : unit-tests/test/script/regex/driver.cxx -*- C++ -*- +// copyright : Copyright (c) 2014-2016 Code Synthesis Ltd +// license : MIT; see accompanying LICENSE file + +#include +#include // is_pod, is_array + +#include + +using namespace std; +using namespace build2::test::script::regex; + +int +main () +{ + using lc = line_char; + using ls = line_string; + using lr = line_regex; + + // Test line_char. + // + { + static_assert (is_pod::value && !is_array::value, + "line_char must be char-like"); + + // Zero-initialed line_char should be the null-char as required by + // char_traits<>::length() specification. + // + assert (lc () == lc::nul); + + line_pool p; + + assert (lc::eof == -1); + assert (lc::nul == 0); + + enum meta {mn = 'n', mp = 'p'}; + + // Special comparison. + // + assert (lc ('0') == lc ('0')); + assert (lc ('0') == '0'); + assert (lc ('n') == mn); + assert (mn == static_cast (lc ('n'))); + + assert (lc ('0') != lc ('1')); + assert (lc ('0') != '1'); + assert (lc ('n') != mp); + assert (lc ('0') != lc ("0", p)); + assert (lc ('0') != lc (regex ("0"), p)); + + assert (lc ('0') < lc ('1')); + assert (lc ('0') < '1'); + assert (lc ('1') < lc ("0", p)); + assert (lc ('n') < mp); + + assert (lc ('0') <= '1'); + assert (lc ('0') <= lc ('1')); + assert (lc ('n') <= mn); + assert (lc ('1') <= lc ("0", p)); + + // Literal comparison. + // + assert (lc ("a", p) == lc ("a", p)); + assert (lc ("a", p).literal == lc ("a", p).literal); + assert (char (lc ("a", p)) == '\a'); + + assert (lc ("a", p) != lc ("b", p)); + assert (!(lc ("a", p) != lc (regex ("a"), p))); // Matches. + assert (lc ("a", p) != lc (regex ("b"), p)); + + assert (lc ("a", p) < lc ("b", p)); + assert (!(lc ("a", p) < lc (regex ("a"), p))); // Matches. + + assert (lc ("a", p) <= lc ("b", p)); + assert (lc ("a", p) <= lc (regex ("a"), p)); + assert (lc ("a", p) < lc (regex ("c"), p)); + + // Regex comparison. + // + assert (lc ("a", p) == lc (regex ("a|b"), p)); + assert (lc (regex ("a|b"), p) == lc ("a", p)); + } + + // Test char_traits. + // + { + using ct = char_traits; + using vc = vector; + + lc c; + ct::assign (c, '0'); + assert (c == ct::char_type ('0')); + + assert (ct::to_char_type (c) == c); + assert (ct::to_int_type (c) == c); + + assert (ct::eq_int_type (c, c)); + assert (!ct::eq_int_type (c, lc::eof)); + + assert (ct::eof () == lc::eof); + + assert (ct::not_eof (c) == c); + assert (ct::not_eof (lc::eof) != lc::eof); + + ct::assign (&c, 1, '1'); + assert (c == ct::int_type ('1')); + + assert (ct::eq (lc ('0'), lc ('0'))); + assert (ct::lt (lc ('0'), lc ('1'))); + + vc v1 ({'0', '1', '2'}); + vc v2 (3, lc::nul); + + assert (ct::find (v1.data (), 3, '1') == v1.data () + 1); + + ct::copy (v2.data (), v1.data (), 3); + assert (v2 == v1); + + v2.push_back (lc::nul); + assert (ct::length (v2.data ()) == 3); + + // Overlaping ranges. + // + ct::move (v1.data () + 1, v1.data (), 2); + assert (v1 == vc ({'0', '0', '1'})); + + v1 = vc ({'0', '1', '2'}); + ct::move (v1.data (), v1.data () + 1, 2); + assert (v1 == vc ({'1', '2', '2'})); + } + + // Test line_char_locale and ctype (only non-trivial functions). + // + { + using ct = ctype; + + line_char_locale l; + assert (has_facet (l)); + + // It is better not to create q facet on stack as it is + // reference-countable. + // + const ct& t (use_facet (l)); + line_pool p; + + assert (t.is (ct::digit, '0')); + assert (!t.is (ct::digit, '?')); + assert (!t.is (ct::digit, lc ("0", p))); + + const lc chars[] = { '0', '?' }; + ct::mask m[2]; + + const lc* b (chars); + const lc* e (chars + 2); + + // Cast flag value to mask type and compare to mask. + // + auto fl = [] (ct::mask m, ct::mask f) {return m == f;}; + + t.is (b, e, m); + assert (fl (m[0], ct::digit) && fl (m[1], 0)); + + assert (t.scan_is (ct::digit, b, e) == b); + assert (t.scan_is (0, b, e) == b + 1); + + assert (t.scan_not (ct::digit, b, e) == b + 1); + assert (t.scan_not (0, b, e) == b); + + { + char nr[] = "0?"; + lc wd[2]; + t.widen (nr, nr + 2, wd); + assert (wd[0] == b[0] && wd[1] == b[1]); + } + + { + lc wd[] = {'0', lc ("a", p)}; + char nr[2]; + t.narrow (wd, wd + 2, '-', nr); + assert (nr[0] == '0' && nr[1] == '-'); + } + } + + // Test regex_traits. Functions other that value() are trivial. + // + { + regex_traits t; + + const int radix[] = {8, 10}; // Radix 16 is not supported by line_char. + const char digits[] = "0123456789ABCDEF"; + + for (size_t r (0); r < 2; ++r) + { + for (int i (0); i < radix[r]; ++i) + assert (t.value (digits[i], radix[r]) == i); + } + } + + // Test line_regex construction. + // + { + line_pool p; + lr r1 ({lc ("foo", p), lc (regex ("ba(r|z)"), p)}, move (p)); + + lr r2 (move (r1)); + assert (regex_match (ls ({lc ("foo", r2.pool), lc ("bar", r2.pool)}), r2)); + assert (!regex_match (ls ({lc ("foo", r2.pool), lc ("ba", r2.pool)}), r2)); + } + + // Test line_regex match. + // + { + line_pool p; + + const lc foo ("foo", p); + const lc bar ("bar", p); + const lc baz ("baz", p); + const lc blank ("", p); + + assert (regex_match (ls ({foo, bar}), lr ({foo, bar}))); + assert (!regex_match (ls ({foo, baz}), lr ({foo, bar}))); + + assert (regex_match (ls ({bar, foo}), + lr ({'(', foo, '|', bar, ')', '+'}))); + + assert (regex_match (ls ({foo, foo}), lr ({'(', foo, ')', '\\', '1'}))); + + assert (regex_match (ls ({foo}), lr ({lc (regex ("fo+"), p)}))); + assert (regex_match (ls ({foo}), lr ({lc (regex (".*"), p)}))); + assert (regex_match (ls ({blank}), lr ({lc (regex (".*"), p)}))); + + assert (regex_match (ls ({blank, blank, foo}), + lr ({blank, '*', foo, blank, '*'}))); + + assert (regex_match (ls ({blank, blank, foo}), lr ({'.', '*'}))); + + assert (regex_match (ls ({blank, blank}), + lr ({blank, '*', foo, '?', blank, '*'}))); + + assert (regex_match (ls ({foo}), lr ({foo, '{', '1', '}'}))); + assert (regex_match (ls ({foo, foo}), lr ({foo, '{', '1', ',', '}'}))); + + assert (regex_match (ls ({foo, foo}), + lr ({foo, '{', '1', ',', '2', '}'}))); + + assert (!regex_match (ls ({foo, foo}), + lr ({foo, '{', '3', ',', '4', '}'}))); + + assert (regex_match (ls ({foo}), lr ({'(', '?', '=', foo, ')', foo}))); + assert (regex_match (ls ({foo}), lr ({'(', '?', '!', bar, ')', foo}))); + } +} -- cgit v1.1