20691103835

Committed 04 Jan 2026 09:49AM UTC coverage: 64.291% (+0.001%) from 64.29%

Build # 20691103835

Build Type

push

github

Committed by

web-flow

Commit Message

1141 improve unified diff (#1142)

* improve unified_diff

* initialize members

* simplified parse

Run Details

18529 of 31593 branches covered (58.65%)

Branch coverage included in aggregate %.

14754 of 20176 relevant lines covered (73.13%)

1515.13 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

91.74

/src/factory/unified-diff-parser.cpp

////////////////////////////////////////////////////////////////////////////////
// Name:      unified-diff-parser.cpp
// Purpose:   Implementation of unified_diff_parser
//            https://www.gnu.org/software/diffutils/manual/html_node/Detailed-Unified.html
// Author:    Anton van Wezenbeek
// Copyright: (c) 2025-2026 Anton van Wezenbeek
////////////////////////////////////////////////////////////////////////////////

#include <boost/parser/parser.hpp>
#include <wex/factory/unified-diff.h>

#include "unified-diff-parser.h"

namespace bp = boost::parser;

wex::factory::unified_diff_parser::unified_diff_parser(unified_diff* diff)
  : m_diff(diff)
{
  m_diff->m_range.fill({0});
  m_diff->m_diffs    = 0;
  m_diff->m_is_first = true;
  m_diff->m_is_last  = true;
  m_diff->m_type     = unified_diff::diff_t::UNKNOWN;
}

bool wex::factory::unified_diff_parser::parse()
{
  auto const action_diff = [this](const auto& ctx)
  {
    const auto tpl    = _attr(ctx);
    m_diff->m_path[0] = wex::path(std::get<0>(tpl));
    m_diff->m_path[1] = wex::path(std::get<1>(tpl));
    m_diff->m_range.fill({0});

    for (const auto& hunk : std::get<2>(tpl))
    {
      int index = 0;

      m_diff->m_is_first = (hunk == *std::get<2>(tpl).begin());
      m_diff->m_is_last  = (hunk == std::get<2>(tpl).back());

      for (const auto& number : std::get<0>(hunk))
      {
        if (const auto* val = std::get_if<std::tuple<int, int>>(&number); val)
        {
          const int range            = std::get<1>(*val);
          m_diff->m_range[index]     = std::abs(std::get<0>(*val));
          m_diff->m_range[index + 1] = range;

          if (range > 0)
          {
            m_diff->m_diffs++;
          }
        }
        else
        {
          m_diff->m_range[index]     = std::abs(std::get<int>(number));
          m_diff->m_range[index + 1] = 1;
          m_diff->m_diffs++;
        }

        index += 2;
      }

      m_diff->m_text.fill({});

      for (const auto& line : std::get<1>(hunk))
      {
        auto fix(line);

        if (line.starts_with("\n"))
        {
          fix = fix.substr(1);
        }

        switch (fix[0])
        {
          case '+':
            m_diff->m_text[1].push_back(fix.substr(1));
            break;
          case '-':
            m_diff->m_text[0].push_back(fix.substr(1));
            break;
          case ' ':
            m_diff->m_text[0].push_back(fix.substr(1));
            m_diff->m_text[1].push_back(fix.substr(1));
            break;
        }
      }

      m_diff->m_type =
        (m_diff->m_type == unified_diff::diff_t::UNKNOWN ?
           unified_diff::diff_t::FIRST :
           unified_diff::diff_t::OTHER);

      m_diff->report_diff();
      m_diff->trace("found");
    }
  };

  auto const action_eoi = [this](const auto& ctx)
  {
    m_diff->m_type = unified_diff::diff_t::LAST;
    m_diff->report_diff_finish();
    m_diff->trace("finish");
  };

  // (Skip the first lines)
  // The unified output format starts with a two-line header:
  // --- from-file from-file-modification-time
  // +++ to-file to-file-modification-time
  // Next come one or more hunks of differences:
  // @@ from-file-line-numbers to-file-line-numbers @@
  // line-from-either-file
  // line-from-either-file...

  auto const parser_diff_lines = bp::lexeme[+(
    bp::char_ >> +(bp::char_ - bp::eol - "--- a/" - "@@" - "diff --"))];

  auto const parser_hunk =
    bp::lit("@@") >> bp::repeat(2)[bp::int_ >> ',' >> bp::int_ | bp::int_] >>
    bp::lit("@@") >> bp::lexeme[+(bp::char_ - bp::eol)] >> parser_diff_lines;

  auto const parser_diff = bp::lit("--- a/") >> +(bp::char_ - "+++ b/") >>
                           bp::lit("+++ b/") >> +(bp::char_ - "@@") >>
                           +parser_hunk;

  auto const parser_skip = bp::omit[*(bp::char_ - "--- a/")];

  auto const parser_all =
    *(parser_skip >> +parser_diff[action_diff]) >> bp::eoi[action_eoi];

  return bp::parse(m_diff->input(), parser_all, bp::ws);
}

1	////////////////////////////////////////////////////////////////////////////////
2	// Name: unified-diff-parser.cpp
3	// Purpose: Implementation of unified_diff_parser
4	// https://www.gnu.org/software/diffutils/manual/html_node/Detailed-Unified.html
5	// Author: Anton van Wezenbeek
6	// Copyright: (c) 2025-2026 Anton van Wezenbeek
7	////////////////////////////////////////////////////////////////////////////////
8
9	#include <boost/parser/parser.hpp>
10	#include <wex/factory/unified-diff.h>
11
12	#include "unified-diff-parser.h"
13
14	namespace bp = boost::parser;
15
16	wex::factory::unified_diff_parser::unified_diff_parser(unified_diff* diff)	14✔
17	: m_diff(diff)	14✔
18	{
19	m_diff->m_range.fill({0});	14✔
20	m_diff->m_diffs = 0;	14✔
21	m_diff->m_is_first = true;	14✔
22	m_diff->m_is_last = true;	14✔
23	m_diff->m_type = unified_diff::diff_t::UNKNOWN;	14✔
24	}	14✔
25
26	bool wex::factory::unified_diff_parser::parse()	14✔
27	{
28	auto const action_diff = [this](const auto& ctx)	14✔
29	{
30	const auto tpl = _attr(ctx);	14✔
31	m_diff->m_path[0] = wex::path(std::get<0>(tpl));	14✔
32	m_diff->m_path[1] = wex::path(std::get<1>(tpl));	14✔
33	m_diff->m_range.fill({0});	14✔
34
35	for (const auto& hunk : std::get<2>(tpl))	34✔
36	{
37	int index = 0;	20✔
38
39	m_diff->m_is_first = (hunk == *std::get<2>(tpl).begin());	20✔
40	m_diff->m_is_last = (hunk == std::get<2>(tpl).back());	20✔
41
42	for (const auto& number : std::get<0>(hunk))	60✔
43	{
44	if (const auto* val = std::get_if<std::tuple<int, int>>(&number); val)	40✔
45	{
46	const int range = std::get<1>(*val);	14✔
47	m_diff->m_range[index] = std::abs(std::get<0>(*val));	14✔
48	m_diff->m_range[index + 1] = range;	14✔
49
50	if (range > 0)	14✔
51	{
52	m_diff->m_diffs++;	4✔
53	}
54	}
55	else
56	{
57	m_diff->m_range[index] = std::abs(std::get<int>(number));	26✔
58	m_diff->m_range[index + 1] = 1;	26✔
59	m_diff->m_diffs++;	26✔
60	}
61
62	index += 2;	40✔
63	}
64
65	m_diff->m_text.fill({});	20✔
66
67	for (const auto& line : std::get<1>(hunk))	68✔
68	{
69	auto fix(line);	48✔
70
71	if (line.starts_with("\n"))	48✔
72	{
73	fix = fix.substr(1);	8✔
74	}
75
76	switch (fix[0])	48!
77	{
78	case '+':	13✔
79	m_diff->m_text[1].push_back(fix.substr(1));	13✔
80	break;	13✔
81	case '-':	20✔
82	m_diff->m_text[0].push_back(fix.substr(1));	20✔
83	break;	20✔
84	case ' ':	×
85	m_diff->m_text[0].push_back(fix.substr(1));	×
86	m_diff->m_text[1].push_back(fix.substr(1));	×
87	break;	×
88	}
89	}
90
91	m_diff->m_type =	20✔
92	(m_diff->m_type == unified_diff::diff_t::UNKNOWN ?	20✔
93	unified_diff::diff_t::FIRST :
94	unified_diff::diff_t::OTHER);
95
96	m_diff->report_diff();	20✔
97	m_diff->trace("found");	60✔
98	}
99	};	28✔
100
101	auto const action_eoi = [this](const auto& ctx)	10✔
102	{
103	m_diff->m_type = unified_diff::diff_t::LAST;	10✔
104	m_diff->report_diff_finish();	10✔
105	m_diff->trace("finish");	18✔
106	};	23✔
107
108	// (Skip the first lines)
109	// The unified output format starts with a two-line header:
110	// --- from-file from-file-modification-time
111	// +++ to-file to-file-modification-time
112	// Next come one or more hunks of differences:
113	// @@ from-file-line-numbers to-file-line-numbers @@
114	// line-from-either-file
115	// line-from-either-file...
116
117	auto const parser_diff_lines = bp::lexeme[+(	14✔
118	bp::char_ >> +(bp::char_ - bp::eol - "--- a/" - "@@" - "diff --"))];
119
120	auto const parser_hunk =
121	bp::lit("@@") >> bp::repeat(2)[bp::int_ >> ',' >> bp::int_ \| bp::int_] >>	14✔
122	bp::lit("@@") >> bp::lexeme[+(bp::char_ - bp::eol)] >> parser_diff_lines;	14✔
123
124	auto const parser_diff = bp::lit("--- a/") >> +(bp::char_ - "+++ b/") >>	14✔
125	bp::lit("+++ b/") >> +(bp::char_ - "@@") >>	14✔
126	+parser_hunk;	14✔
127
128	auto const parser_skip = bp::omit[*(bp::char_ - "--- a/")];	14✔
129
130	auto const parser_all =
131	*(parser_skip >> +parser_diff[action_diff]) >> bp::eoi[action_eoi];	14✔
132
133	return bp::parse(m_diff->input(), parser_all, bp::ws);	27✔
134	}

antonvw / wex / 20691103835

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous