Kcov zig 4merge (for discussion)

.vscode/launch.json

@@ -0,0 +1,21 @@
+{
+  // Use IntelliSense to learn about possible attributes.
+  // Hover to view descriptions of existing attributes.
+  // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+  "version": "0.2.0",
+  "configurations": [
+    {
+      "type": "lldb",
+      "request": "launch",
+      "name": "Debug",
+      "program": "${workspaceFolder}/build/src/kcov",
+      "args": [
+        "./cov",
+        "--include-path=/Users/bytedance/DevCamp/jstring",
+        "--clean",
+        "/Users/bytedance/DevCamp/jstring/zig-out/bin/pcre_test"
+      ],
+      "cwd": "${workspaceFolder}/../jstring"
+    }
+  ]
+}

README.md

@@ -1,13 +1,62 @@
+# A fork of `kcov` for better covering Zig
+
+This is a rather than naive fork of wonderful `kcov` for better supporting Zig projects.
+
+## It adds the ability to auto recognize Zig's `unreachable` and `@panic` for auto ignoring.
+
+Here is a screenshot of auto ignoring `unreachable`.
+
+![zig-unreachable-example](https://github.com/liyu1981/kcov/blob/master/nocover/zig-unreachable.png?raw=true)
+
+Here is another screenshot of auto ignoring `@panic`.
+
+![zig-panic-example](https://github.com/liyu1981/kcov/blob/master/nocover/zig-panic.png?raw=true)
+
+## It also adds the ability in c/c++ source file of using `/* no-cover */` to mark a line to be ignored.
+
+Here is a screenshot of manual ignoring by placing `/* no-cover */` in c source file.
+
+![c-no-cover example](https://github.com/liyu1981/kcov/blob/master/nocover/c-nocover.png?raw=true)
+
+## Usage
+
+There is no change of original `kcov` usage, it will just work. Please follow the below original Kcov README.
+
+### But how can I get the binary?
+
+The best way is to compile from source. It can be done as follows (you will need `cmake`, `ninja`, `llvm@>16` as I tried)
+
+```
+git clone https://github.com/liyu1981/kcov.git
+cd kcov
+mkdir build
+cd build
+CC="clang" CXX="clang++" cmake -G Ninja ..
+ninja
+```
+
+after building is done. The binary is at `build/src/kcov`. Copy somewhere and use it.
+
+(Or you can download a copy of Apple Silicon version binary from the release section.)
+
+## File Support
+
+- for Zig, support source file with `.zig` extension.
+- for C/C++, support source file with `.c/.cpp/.cc` extension.
+
+## Original Kcov Readme
+
 [![Coveralls coverage status](https://img.shields.io/coveralls/SimonKagstrom/kcov.svg)](https://coveralls.io/r/SimonKagstrom/kcov?branch=master)
 [![Codecov coverage status](https://codecov.io/gh/SimonKagstrom/kcov/branch/master/graph/badge.svg)](https://codecov.io/gh/SimonKagstrom/kcov)
 [![Coverity Scan Build Status](https://scan.coverity.com/projects/2844/badge.svg)](https://scan.coverity.com/projects/2844)
 ![Docker Pulls](https://img.shields.io/docker/pulls/kcov/kcov.svg)
 
 [![PayPal Donate](https://img.shields.io/badge/paypal-donate-blue.svg)](https://www.paypal.com/cgi-bin/webscr?cmd=_donations&business=simon.kagstrom%40gmail%2ecom&lc=US&item_name=Simon%20Kagstrom&item_number=kcov&currency_code=USD&bn=PP%2dDonationsBF%3abtn_donate_LG%2egif%3aNonHosted) [![Github All Releases](https://img.shields.io/github/downloads/atom/atom/total.svg)](https://github.com/SimonKagstrom/kcov/)
 
-## *kcov*
+## _kcov_
+
 Kcov is a FreeBSD/Linux/OSX code coverage tester for compiled languages, Python
-and Bash.  Kcov was originally a fork of [Bcov](http://bcov.sf.net), but has
+and Bash. Kcov was originally a fork of [Bcov](http://bcov.sf.net), but has
 since evolved to support a large feature set in addition to that of Bcov.
 
 Kcov, like Bcov, uses DWARF debugging information for compiled programs to
@@ -16,28 +65,27 @@ switches.
 
 For a video introduction, [look at this presentation from SwedenCPP](https://www.youtube.com/watch?v=1QMHbp5LUKg)
 
-Installing
-----------
+## Installing
+
 Refer to the [INSTALL](INSTALL.md) file for build instructions, or use our official Docker images:
 
-* [kcov/kcov](https://hub.docker.com/r/kcov/kcov/) for releases since v31.
+- [kcov/kcov](https://hub.docker.com/r/kcov/kcov/) for releases since v31.
 
+## How to use it
 
-How to use it
--------------
 Basic usage is straight-forward:
 
 ```
 kcov /path/to/outdir executable [args for the executable]
 ```
 
-*/path/to/outdir* will contain lcov-style HTML output generated
+_/path/to/outdir_ will contain lcov-style HTML output generated
 continuously while the application runs. Kcov will also write cobertura-
 compatible XML output and generic JSON coverage information and can easily
 be integrated in various CI systems.
 
-Filtering output
-----------------
+## Filtering output
+
 It's often useful to filter output, since e.g., /usr/include is seldom of interest.
 This can be done in two ways:
 
@@ -47,8 +95,8 @@ kcov --exclude-pattern=/usr/include --include-pattern=part/of/path,other/path \
 ```
 
 which will do a string-comparison and include everything which contains
-*part/of/path* or *other/path* but exclude everything that has the
-*/usr/include* string in it.
+_part/of/path_ or _other/path_ but exclude everything that has the
+_/usr/include_ string in it.
 
 ```
 kcov --include-path=/my/src/path /path/to/outdir executable
@@ -57,8 +105,8 @@ kcov --exclude-path=/usr/include /path/to/outdir executable
 
 Does the same thing, but with proper path lookups.
 
-Merging multiple kcov runs
---------------------------
+## Merging multiple kcov runs
+
 Kcov can also merge the results of multiple earlier runs. To use this mode,
 call kcov with `--merge`, an output path and one or more paths to an earlier
 run, e.g.,
@@ -68,20 +116,20 @@ kcov --merge /tmp/merged-output /tmp/kcov-output1 /tmp/kcov-output2
 kcov --merge /tmp/merged-output /tmp/kcov-output*    # With a wildcard
 ```
 
-Use from continuous integration systems
----------------------------------------
+## Use from continuous integration systems
+
 kcov is easy to integrate with [travis-ci](http://travis-ci.org) together with
 [coveralls.io](http://coveralls.io) or [codecov.io](http://codecov.io). It can also
 be used from Jenkins, [SonarQube](http://sonarqube.org) and [GitLab CI](http://gitlab.com).
 Refer to
 
-* [coveralls](doc/coveralls.md) for details about travis-ci + coveralls, or
-* [codecov](doc/codecov.md) for details about travis-ci + codecov
-* [jenkins](doc/jenkins.md) for details about how to integrate in Jenkins
-* [sonarqube](doc/sonarqube.md) for how to use kcov and sonarqube together
-* [gitlab](doc/gitlab.md) for use with GitLab
+- [coveralls](doc/coveralls.md) for details about travis-ci + coveralls, or
+- [codecov](doc/codecov.md) for details about travis-ci + codecov
+- [jenkins](doc/jenkins.md) for details about how to integrate in Jenkins
+- [sonarqube](doc/sonarqube.md) for how to use kcov and sonarqube together
+- [gitlab](doc/gitlab.md) for use with GitLab
+
+## More information
 
-More information
-----------------
 kcov is written by Simon Kagstrom <[email protected]> and more
 information can be found at [the web page](http://simonkagstrom.github.io/kcov/index.html)

src/CMakeLists.txt

@@ -239,6 +239,7 @@ set (${KCOV}_SRCS
     writers/html-writer.cc
     writers/sonarqube-xml-writer.cc
     writers/writer-base.cc
+    writers/nocover.cc
     ${ELF_SRCS}
     ${MACHO_SRCS}
     include/capabilities.hh

src/writers/codecov-writer.cc

@@ -18,6 +18,7 @@ class type_info;
 #include <sstream>
 
 #include "writer-base.hh"
+#include "nocover.hh"
 
 namespace kcov
 {
@@ -134,6 +135,12 @@ class CodecovWriter : public WriterBase
 				IReporter::LineExecutionCount cnt = m_reporter.getLineExecutionCount(file->m_name, n);
 				std::string hitScore = "0";
 
+				std::string& line_str = file->m_lineMap.at(n);
+				std::string& file_name = file->m_fileName;
+				if (!shouldCover(line_str, file_name)) {
+					continue;
+				}
+
 				if (m_maxPossibleHits == IFileParser::HITS_UNLIMITED || m_maxPossibleHits == IFileParser::HITS_SINGLE)
 				{
 					if (cnt.m_hits) {
@@ -167,7 +174,7 @@ class CodecovWriter : public WriterBase
 		linesBlock += "\n";
 
 
-		std::string out = 
+		std::string out =
 			"    \"" + filename + "\": {\n" +
 			linesBlock +
 			"    }";

src/writers/html-writer.cc

@@ -16,6 +16,7 @@
 #include <fstream>
 
 #include "writer-base.hh"
+#include "nocover.hh"
 
 using namespace kcov;
 
@@ -72,6 +73,11 @@ class HtmlWriter : public WriterBase
 					"\"line\":\"", n);
 			outJson << escape_json(line) << "\"";
 
+			std::string& line_str = file->m_lineMap[n];
+			std::string& file_name = file->m_name;
+			bool should_cover = shouldCover(line_str, file_name);
+			const std::string& no_cover_class = "lineNoCov";
+
 			if (m_reporter.lineIsCode(file->m_name, n))
 			{
 				IReporter::LineExecutionCount cnt = m_reporter.getLineExecutionCount(file->m_name, n);
@@ -90,18 +96,21 @@ class HtmlWriter : public WriterBase
 						lineClass = "linePartCov";
 				}
 
-				outJson << fmt(",\"class\":\"%s\","
-						"\"hits\":\"%u\",", lineClass.c_str(), cnt.m_hits);
-
-				// Don't report order for zeroes
-				if (cnt.m_order)
-					outJson << fmt("\"order\":\"%llu\",", (unsigned long long) cnt.m_order);
+				if (lineClass == no_cover_class && !should_cover) {
+					// should not cover, so skip, but if there are some hits (when lineClass is not no cov), go else for normal marking
+				} else {
+					outJson << fmt(",\"class\":\"%s\","
+							"\"hits\":\"%u\",", lineClass.c_str(), cnt.m_hits);
+					// Don't report order for zeroes
+					if (cnt.m_order)
+						outJson << fmt("\"order\":\"%llu\",", (unsigned long long) cnt.m_order);
 
-				if (m_maxPossibleHits != IFileParser::HITS_SINGLE)
-					outJson << fmt("\"possible_hits\":\"%u\",", cnt.m_possibleHits);
+					if (m_maxPossibleHits != IFileParser::HITS_SINGLE)
+						outJson << fmt("\"possible_hits\":\"%u\",", cnt.m_possibleHits);
 
-				nExecutedLines += !!cnt.m_hits;
-				nCodeLines++;
+					nExecutedLines += !!cnt.m_hits;
+					nCodeLines++;
+				}
 			}
 			outJson << "},\n";
 
@@ -252,15 +261,29 @@ class HtmlWriter : public WriterBase
 			if (!res)
 				continue;
 
+			// since we may ignored some from previous individual writeOne functions for each file, here use the sum from
+			// files instead of summary
+			int inFilesTotalCodeLines = 0;
+			int inFilesTotalExecutedLines = 0;
+			for (FileMap_t::const_iterator it = m_files.begin(); it != m_files.end(); ++it)
+			{
+				inFilesTotalCodeLines += it->second->m_codeLines;
+				inFilesTotalExecutedLines += it->second->m_executedLines;
+			}
+
 			// Skip entries (merged ones) that shouldn't be included in the totals
 			if (summary.m_includeInTotals)
 			{
-				nTotalCodeLines += summary.m_lines;
-				nTotalExecutedLines += summary.m_executedLines;
+					nTotalCodeLines += inFilesTotalCodeLines;
+					nTotalExecutedLines += inFilesTotalExecutedLines;
+					// nTotalCodeLines += summary.m_lines;
+					// nTotalExecutedLines += summary.m_executedLines;
 			}
 
-			std::string datum = getIndexHeader(fmt("%s/index.html", de->d_name), name, name, summary.m_lines,
-					summary.m_executedLines);
+			std::string datum = getIndexHeader(fmt("%s/index.html", de->d_name), name, name, inFilesTotalCodeLines,
+					inFilesTotalExecutedLines);
+			// std::string datum = getIndexHeader(fmt("%s/index.html", de->d_name), name, name, summary.m_lines,
+			// summary.m_executedLines);
 
 			if (name == conf.keyAsString("merged-name"))
 				merged += datum;

src/writers/json-writer.cc

@@ -16,6 +16,7 @@ class type_info;
 #include <fstream>
 
 #include "writer-base.hh"
+#include "nocover.hh"
 
 using namespace kcov;
 
@@ -61,12 +62,21 @@ class JsonWriter : public WriterBase
 			for (unsigned int n = 1; n < file->m_lastLineNr; n++)
 			{
 				IReporter::LineExecutionCount cnt = m_reporter.getLineExecutionCount(file->m_name, n);
+
+				const std::string& line_str = file->m_lineMap[n];
+				const std::string& file_name = it->first;
+				bool should_cover = shouldCover(line_str, file_name);
+
 				if (m_reporter.lineIsCode(file->m_name, n))
 				{
-					nExecutedLines += !!cnt.m_hits;
-					nCodeLines++;
-					nTotalExecutedLines += !!cnt.m_hits;
-					nTotalCodeLines++;
+					if (should_cover) {
+						nExecutedLines += !!cnt.m_hits;
+						nCodeLines++;
+						nTotalExecutedLines += !!cnt.m_hits;
+						nTotalCodeLines++;
+					} else {
+						// skip this from counting
+					}
 				}
 			}
 			if (nCodeLines > 0)