README (4348B)
1 stagit 2 ====== 3 4 WARNING: The official stagit page is there: https://git.2f30.org/stagit/ 5 This is just a mirror for my own use, you probably don't want to use this! 6 7 static git page generator 8 9 10 Usage 11 ----- 12 13 Make files per repository: 14 15 $ mkdir -p htmldir && cd htmldir 16 $ stagit path-to-repo 17 18 Make index file for repositories: 19 20 $ stagit-index repodir1 repodir2 repodir3 > index.html 21 22 23 Install 24 ------- 25 26 $ make 27 # make install 28 29 30 Dependencies 31 ------------ 32 33 - libgit2 (v0.22+). 34 - libc (tested with OpenBSD, FreeBSD, Linux: glibc and musl). 35 - C compiler (C99). 36 - make 37 38 39 Documentation 40 ------------- 41 42 See man pages: stagit(1) and stagit-index(1). 43 44 45 Building a static binary 46 ------------------------ 47 48 It may be useful to build static binaries, for example to run in a chroot. 49 50 It can be done like this at the time of writing (v0.24): 51 52 cd libgit2-src 53 54 # change the options in the CMake file: CMakeLists.txt 55 BUILD_SHARED_LIBS to OFF (static) 56 CURL to OFF (not needed) 57 USE_SSH OFF (not needed) 58 THREADSAFE OFF (not needed) 59 USE_OPENSSL OFF (not needed, use builtin) 60 61 mkdir -p build && cd build 62 cmake ../ 63 make 64 make install 65 66 67 Extract owner field from git config 68 ----------------------------------- 69 70 A way to extract the gitweb owner for example in the format: 71 72 [gitweb] 73 owner = Name here 74 75 Script: 76 77 #!/bin/sh 78 awk '/^[ ]*owner[ ]=/ { 79 sub(/^[^=]*=[ ]*/, ""); 80 print $0; 81 }' 82 83 84 Set clone url for a directory of repos 85 -------------------------------------- 86 #!/bin/sh 87 cd "$dir" 88 for i in *; do 89 test -d "$i" && echo "git://git.codemadness.org/$i" > "$i/url" 90 done 91 92 93 Update files on git push 94 ------------------------ 95 96 Using a post-receive hook the static files can be automatically updated. 97 Keep in mind git push -f can change the history and the commits may need 98 to be recreated. This is because stagit checks if a commit file already 99 exists. It also has a cache (-c) option which can conflict with the new 100 history. See stagit(1). 101 102 git post-receive hook (repo/.git/hooks/post-receive): 103 104 #!/bin/sh 105 # detect git push -f 106 force=0 107 while read -r old new ref; do 108 hasrevs=$(git rev-list "$old" "^$new" | sed 1q) 109 if test -n "$hasrevs"; then 110 force=1 111 break 112 fi 113 done 114 115 # remove commits and .cache on git push -f 116 #if test "$force" = "1"; then 117 # ... 118 #fi 119 120 # see example_create.sh for normal creation of the files. 121 122 123 Create .tar.gz archives by tag 124 ------------------------------ 125 #!/bin/sh 126 name="stagit" 127 mkdir -p archives 128 git tag -l | while read -r t; do 129 f="archives/${name}-$(echo "${t}" | tr '/' '_').tar.gz" 130 test -f "${f}" && continue 131 git archive \ 132 --format tar.gz \ 133 --prefix "${t}/" \ 134 -o "${f}" \ 135 -- \ 136 "${t}" 137 done 138 139 140 Features 141 -------- 142 143 - Log of all commits from HEAD. 144 - Log and diffstat per commit. 145 - Show file tree with linkable line numbers. 146 - Show references: local branches and tags. 147 - Detect README and LICENSE file from HEAD and link it as a webpage. 148 - Detect submodules (.gitmodules file) from HEAD and link it as a webpage. 149 - Atom feed log (atom.xml). 150 - Make index page for multiple repositories with stagit-index. 151 - After generating the pages (relatively slow) serving the files is very fast, 152 simple and requires little resources (because the content is static), only 153 a HTTP file server is required. 154 - Usable with text-browsers such as dillo, links, lynx and w3m. 155 156 157 Cons 158 ---- 159 160 - Not suitable for large repositories (2000+ commits), because diffstats are 161 an expensive operation, the cache (-c flag) is a workaround for this in 162 some cases. 163 - Not suitable for large repositories with many files, because all files are 164 written for each execution of stagit. This is because stagit shows the lines 165 of textfiles and there is no "cache" for file metadata (this would add more 166 complexity to the code). 167 - Not suitable for repositories with many branches, a quite linear history is 168 assumed (from HEAD). 169 170 In these cases it is better to just use cgit or possibly change stagit to 171 run as a CGI program. 172 173 - Relatively slow to run the first time (about 3 seconds for sbase, 174 1500+ commits), incremental updates are faster. 175 - Does not support some of the dynamic features cgit has, like: 176 - Snapshot tarballs per commit. 177 - File tree per commit. 178 - History log of branches diverged from HEAD. 179 - Stats (git shortlog -s). 180 181 This is by design, just use git locally.