#!/bin/bash ## GophHub - Gopher proxy for GitHub API VERSION=1.2 ## Designed for Gophernicus CGI environment ## ## URL parameters: ## # repo=[ GitHub repo "owner/name" ] # L If missing, will extract repo "owner/name" and directory/file path from GitHub URL in $SEARCHREQUEST # # path=[ full directory/file path within repo ] # # file=1 # L Retrieve file instead of directory listing # file=readme # L Retrieve HTML-formatted default README file for the repo # # raw=1 # L Retrieve raw files # # html=1 # L Retrieve HTML-formatted Markdown/ASCIIDoc files, or attempt syntax highlighting with Enscript # # num=1 # L Add line numbers to source code files # # settings=1 # L Display settings menu ## ## Configuration Options: ## # Path to store links to programs allowed to run in Bash restricted mode EXECPATH=/home/freet/public_gopher/gophhub/bin # Gophermap to display as "homepage", when no repo is specified MENUMAP=/home/freet/public_gopher/gophhub/menu.gophermap # CGI script location CGIPATH='/~freet/cgi-bin/gophhub.sh' # Gophermap location GOPHERMAPPATH='/~freet/gophhub/' # API result cache location (directory will be created if it doesn't exist) CACHEDIR=/tmp/gophhub # API result cache time (minutes) CACHETIME=15 # Maximum file size (in bytes) to offer API file download with MAXSIZE=512000 # Maximum number of lines from the README text to be shown in repo root directory gophermap READMEMAX=25 # Maximum width of wrapped text in repo root gophermap ($COLUMNS value is set by Gophernicus) WRAP=$COLUMNS # GitHub API token (optional) https://docs.github.com/en/github/authenticating-to-github/creating-a-personal-access-token APITOKEN= # ## BEGIN SCRIPT ## [ $WRAP -gt 2 ] && WRAP=$[ $WRAP - 1 ] || WRAP=66 # Set up GitHub API request headers VHEADER='X-GitHub-Api-Version: 2022-11-28' [ "$APITOKEN" ] && AUTHHEADER="Authorization: Bearer $APITOKEN" || AUTHHEADER= # Display HTML-rendered text content htmlview () { fullpath="$repo/$path" if [ "$1" == "readme.htm" -o "$1" == "mdfile.htm" ] then echo -e '\n'\ '\n\n'\ " GophHub - $repo README\n"\ ' \n'\ '\n\n' if [ "$1" == "readme.htm" ] then echo "

GophHub - $repo README

" else # Remove file name from path [ "${fullpath%/*}" == "$repo" ] && urlpath= || urlpath="${path%/*}" echo -e "

GophHub - ${fullpath%/}/${path##/}

\n"\ "

Raw File

" fi # HTML-viewable file format extensions regexp # Markdown & ASCIIDoc extensions + source code formats supported by Enscript (in enscript-1.6.6/states/hl/enscript.st) viewable='$md$\?$markdown\?$\?$adoc$\?$asciidoc$\?$txt$\?$c$\?$h$\?$c++$\?$cpp$\?$cc$\?$cxx$\?$m$\?$mpl$\?$mp$\?$maple$\?$scm$\?$scheme$\?$ads$\?$adb$\?$ada$\?$s$\?$st$\?$MOD$\?$DEF$\?$mi$\?$tcl$\?$v$\?$vh$\?$html$\?$vhd$\?$vhdl$\?$scr$\?$syn$\?$synth$\?$idl$\?$hs$\?$lhs$\?$gs$\?$lgs$\?$pm$\?$pl$\?$eps$\?$ps$\?$py$\?$pyx$\?$js$\?$java$\?$pas$\?$pp$\?$p$\?$f$\?$f90$\?$awk$\?$sh$\?$vba$\?$csh$\?$m4$\?$il$\?$wrl$\?$inf$\?$tex$\?$wmlscript$\?$wmls$\?' # Remove file name from path (adds leading slash for ascending from first sub-directory in step 5 below) [ "${fullpath%/*}" == "$repo" ] && urlpath= || urlpath=/"${path%/*}" # Convert GitHub links in HTML to GophHub/raw-download links (needs GNU Sed extensions): # 1) Replace absolute links to root directories of GitHub repos with GophHub links # 2) Replace absolute links to sub-directories of GitHub repos with GophHub links # 3) Replace absolute links to MarkDown or ASCIIdoc files with HTML-rendered GophHub links # 4) Replace absolute links to files at GitHub with raw.githubusercontent.com URLs # 5) Replace relative links to files with absolute raw.githubusercontent.com URLs ("../" to parent) # 6) Replace relative links to files with absolute raw.githubusercontent.com URLs (lower directories only) # 7) Convert relative directory links from 5 and 6 to GophHub links # 8) Convert relative MarkDown or ASCIIdoc file links to HTML-rendered GophHub links # 9) Fix "id" elements by removing "user-content-", which is prepended by GitHub's HTML renderer sed -e 's%href="https*://w*\.*github\.com/$[^/"]*/[^/#"]*$/\?[^/"]*"%href="'"gopher://$SERVER_NAME/1$GOPHERMAPPATH?repo=\1\&html=1\&num=$num"'"%ig' \ -e 's%href="https*://w*\.*github\.com/$[^/"]*/[^/"]*$/tree/[^/"]*/$[^#"]*$"%href="'"gopher://$SERVER_NAME/1$GOPHERMAPPATH?repo=\1\&path=\2\&html=1\&num=$num"'"%ig' \ -e 's%href="https*://w*\.*github\.com/$[^/"]*/[^/"]*$/blob/[^/"]*/$[^"]*\.'"$viewable"'$$#[^"]*$*"%href="'"gopher://$SERVER_NAME/h$CGIPATH?repo=\1\&path=\2\&html=1\&num=$num\&file=1"'"%ig' \ -e 's%$\(src=$\|$href=$\)"https*://w*\.*github\.com/$[^/"]*/[^/"]*$/$raw$\?$blob$\?/-\?$master$\?/$[^#"]*$$#[^"]*$*"%\1"https://raw.githubusercontent.com/\4/master/\8"%ig' \ -e 's%$\(src=$\|$href=$\)"\.\.\/\?$[^#":]*$$#[^"]*$*"%\1"https://raw.githubusercontent.com/'"${repo}/master${urlpath%/*}/"'\4\"%ig' \ -e 's%$\(src=$\|$href=$\)"\.\?/\?$[^#":]\+$$#[^"]*$*"%\1"https://raw.githubusercontent.com/'"${repo}/master$urlpath/"'\4\"%ig' \ -e 's%href="https://raw.githubusercontent.com/$[^/"]*/[^/"]*$/master/$[^".]*$"%href="'"gopher://$SERVER_NAME/1$GOPHERMAPPATH?repo=\1\&path=\2\&html=1\&num=$num"'"%ig' \ -e 's%href="https://raw.githubusercontent.com/$[^/"]*/[^/"]*$/master/$[^"]*\.'"$viewable"'$"%href="'"gopher://$SERVER_NAME/h$CGIPATH?repo=\1\&path=\2\&html=1\&num=$num\&file=1"'"%ig' \ -e 's%id="user-content-%id="%g' "$1" echo -e '

Generated by GophHub '"$VERSION.

\n\n" else # Add GophHub heading to GNU Enscript output # Remove file name from path [ "${fullpath%/*}" == "$repo" ] && urlpath= || urlpath="${path%/*}" echo -e '\n'\ '\n\n'\ " GophHub - $repo/$path\n"\ '\n\n'\ "

GophHub - ${fullpath%/}/${path##/}

\n"\ "

Raw File

" if [ $num -gt 0 ] then # Add line numbers with pr, negatively offset by the number of HTML lines preceeding the preformatted text pr -Tn -N -10 "$1" | sed -n -e 's/[[:blank:]]*0[[:blank:]]*

//' -e '//,/<\/PRE>/p' 
    else
      # Display file contents without line numbering
      sed -n '//,/<\/PRE>/p' "$1"
    fi
    echo -e "
\nGenerated by GNU Enscript 1.6.6, and"\
            'GophHub '"$VERSION.\n\n"
  fi
}

## Main

repo=
path=
carry=
commits=
raw=0
html=0
num=0
file=0
settings=0

# Turn query string into positional parameters, as if they were arguments
oldIFS="$IFS"
IFS='&'
set -- $QUERY_STRING
IFS="$oldIFS"

# Process query string
while [ "$1" ]
do
  case "$1" in
    repo=*) repo="${1#*=}" ;;
    path=*) path="${1#*=}" ;;
    file=*) file="${1#*=}" ;;
    raw=*) raw="${1#*=}"; carry="${carry}&raw=$raw"  ;;
    html=*) html="${1#*=}"; carry="${carry}&html=$html" ;;
    num=*) num="${1#*=}"; carry="${carry}&num=$num" ;;
    settings=1) settings=1 ;;
  esac
  shift
done

# If no "repo" specified, display home menu gophermap
if [ -z "$repo" ]
then
 # Decode percent-encoded characters (eg. from Web proxies)
 SEARCHREQUEST="$(printf '%b' "${SEARCHREQUEST//%/\\x}")"
 # Extract repo "owner/name" from URL in search request
 if repo="`expr \"$SEARCHREQUEST\" : '.*github\.[[:alpha:]]*/\([^/]*/[^/#]*\)'`"
 then
   # Extract directory/file path from URL in search request
   if ! path="`expr \"$SEARCHREQUEST\" : '.*github\.[[:alpha:]]*/[^/]*/[^/]*/tree/[^/]*/\(.*\)'`"
   then
     if path="`expr \"$SEARCHREQUEST\" : '.*github\.[[:alpha:]]*/[^/]*/[^/]*/blob/[^/]*/\([^#]*\)'`"
     then
       file=1
     fi
   fi
 else
  homepage="`cat \"$MENUMAP\"`"
 fi
fi

# Show settings gophermap
if [ $settings -gt 0 -o -z "$repo" ]
then
  # Restricted mode (only using built-ins)
  PATH=
  set -r

  # Display GophHub home menu gophermap ($MENUMAP) with variable values filled in
  if [ -z "$repo" ]
  then
    # For compatibility, only add ? on the end of the search URL if an option is selected
    [ "$carry" ] && search_carry="?${carry#&}" || search_carry=
    eval "echo \"$homepage\""
  fi
  # Display settings options with [*] toggle boxes
  echo -e "!GophHub Settings\nSelect menu item to toggle...\n"
  if [ "$html" != "1" ]; then html=1;mark=" ";else html=0;mark="*"; fi
  echo -e "1[$mark]View files in HTML via Gopher\t${REQUEST}?repo=${repo}${carry//&html=?/}&html=$html&path=${path}"
  if [ "$raw" != "1" ]; then raw=1;mark=" ";else raw=0;mark="*"; fi
  echo -e "1[$mark]Download raw files via Gopher rather than HTTPS\t${REQUEST}?repo=${repo}${carry//&raw=?/}&raw=$raw&path=${path}"
  if [ "$num" != "1" ]; then num=1;mark=" ";else num=0;mark="*"; fi
  echo -e "1[$mark]Number lines in source files (Gopher downloads only)\t${REQUEST}?repo=${repo}${carry//&num=?/}&num=$num&path=${path}"

  echo -e "\nNote: These settings are only applied to files under ${MAXSIZE}B"
  [ "$repo" ] && echo -e "\n1Return to directory list\t${REQUEST}?repo=${repo}${carry}&path=${path}"
  exit
fi

cachecount=0
# Cache downloaded files in "$CACHEDIR"
if [ -d "$CACHEDIR" ]
then
  #Clean up old downloads
  find "$CACHEDIR" -maxdepth 1 -mindepth 1 -type d -mmin +$CACHETIME -exec rm -rf '{}' +
  cachecount=`ls "$CACHEDIR" | wc -l`
else
  mkdir "$CACHEDIR" || exit 1
fi

# Checksum of repo and path string is used as ID for downloaded/cached files
cachesum="`echo \"$repo/$path\" | md5sum`"
cachesum=${cachesum% -}
# Check for cached files
if [ ! -d "$CACHEDIR"/$cachesum ]
then
  if ! mkdir "$CACHEDIR"/$cachesum
  then
   echo -e "ERROR:\nCouldn't create temporary directory"
   exit 1
  fi
fi

cd "$CACHEDIR"/$cachesum
[ -L git_dir.json ] || ln -s index.html git_dir.json

# Link to all the programs that are used after this
for prog in cat sed wget fold pr grep enscript numfmt
do
  [ -f "$EXECPATH"/$prog ] || ln -sf "`which $prog`" "$EXECPATH"/$prog
done

# Set path to only execute intended programs
PATH="$EXECPATH"

# Disable error output
exec 2>/dev/null

# Switch to Bash restricted mode
set -r

# Download data from GitHub API and apply syntax highlighting to files if required
if [ -z "$path" ]
then
  if [ "$file" == "readme" ]
  then
    # View README in HTML
    wget -q -nc -O readme.htm --header="$VHEADER" --header="$AUTHHEADER" --header="Accept: application/vnd.github.html" \
     "https://api.github.com/repos/$repo/readme"
    if [ -s readme.htm ]
    then
     htmlview readme.htm
    else
     echo "GophHub - README not foundREADME empty or not found"
    fi
    exit
  fi
  # if this is the root directory, grab extra info
  wget -q -nc --header="$VHEADER" --header="$AUTHHEADER" --header="Accept: application/vnd.github.raw" "https://api.github.com/repos/$repo" \
   "https://api.github.com/repos/$repo/contents/" "https://api.github.com/repos/$repo/readme" "https://api.github.com/repos/$repo/releases/latest" \
   "https://api.github.com/repos/$repo/commits/master?per_page=1"
  if [ ! -s "${repo##*/}" ]
  then
    echo "Failed to retrieve data for repo: $repo"
    exit
  fi
  [ -s "master?per_page=1" ] && commits="master?per_page=1"
else
  if [ $file -gt 0 ]
  then
    # Find filename extension
    fileext="${path##*.}"
    # If no filename extension, assume plain text
    [ "$fileext" == "$path" ] && fileext=txt
    if [ $html -gt 0 ]
    then
      if [ "$fileext" == "md" -o "$fileext" == "markdown" -o "$fileext" == "adoc" -o "$fileext" == "asciidoc" -o "$fileext" == "txt" ]
      then
        # Download HTML-rendered Markdown or AsciiDoc
        wget -q -nc -O mdfile.htm --header="$VHEADER" --header="$AUTHHEADER" --header="Accept: application/vnd.github.html" \
         "https://api.github.com/repos/$repo/contents/$path"
        if [ -s mdfile.htm ]
        then
         htmlview mdfile.htm
        else
         echo "GophHub - File not foundFile empty or not found"
        fi
        exit
      else
        # Download raw file via API
        wget -q -nc -O "file.$fileext" --header="$VHEADER" --header="$AUTHHEADER" --header="Accept: application/vnd.github.raw" \
         "https://api.github.com/repos/$repo/contents/$path"
        if [ -s "file.$fileext" ]
        then
         # Add syntax highlighting in HTML
         [ -f file.htm ] || enscript -q -E -w html --color -o file.htm "file.$fileext"
         htmlview file.htm
        else
         echo "GophHub - File not foundFile empty or not found"
        fi
        exit
      fi
     else
      # Download raw file via API
      wget -q -nc -O "file.$fileext" --header="$VHEADER" --header="$AUTHHEADER" --header="Accept: application/vnd.github.raw" \
       "https://api.github.com/repos/$repo/contents/$path"
      if [ -s "file.$fileext" ]
      then
       if [ $num -gt 0 ]
       then
         # Display file contents with line numbering
         pr -Tn "file.$fileext"
       else
         # Display file contents
         cat "file.$fileext"
       fi
      else
       echo "File empty or not found"
      fi
      exit
    fi
  else
    # Download directory data
    wget -q -nc --header="$VHEADER" --header="$AUTHHEADER" --header="Accept: application/vnd.github+json" \
     "https://api.github.com/repos/$repo/contents/${path%/}/"
  fi
fi

# Below here is only for displaying directory listings. File retrieval exits above.

# Display repo info
if [ -z "$path" ]
then
  oldIFS="$IFS"
  IFS=\" # remove quotes from data using IFS
  # Read in lines of JSON data with useful fields and parse them by splitting at the '"' character
  # It's slightly faster to filter the input using Grep
  grep '"description":'$'\n''"homepage":'$'\n''"size":'$'\n''"language":'$'\n''"spdx_id":'$'\n''"date":'$'\n''"message":' "${repo#*/}" $commits \
  | while read -r start FIELD sep VALUE comma || {
        # This part runs after the end of the end of output from Grep
        echo "!GophHub - ${repo}:"
        echo "$DESCVAL" | fold -s -w $WRAP | pr -o 1 -T
        [ "$HOMEVAL" ] && echo -e "h Project's Website\tURL:$HOMEVAL"
        echo -e "h Download Repo ZIP Archive\tURL:https://api.github.com/repos/$repo/zipball"
        echo -e "h GitHub URL: https://github.com/$repo.git\tURL:https://github.com/$repo.git\n"\
          "Repo Size: $SIZEVAL\n"\
          "Programming Language: $LANGVAL\n"\
          "License: $LICENSEVAL"
        [ "$DATEVAL" ] && echo -e " Latest Commit Date: $DATEVAL\n"\
                                  "Latest Commit Message:\n$MESSAGEVAL"
        false
       }
  do
   case "$FIELD" in
    description) DESCVAL="$VALUE" ;;
    homepage) HOMEVAL="$VALUE" ;;
    size) sep="${sep#: *}"; SIZEVAL="`numfmt --to=iec --from=iec \"${sep%,}\"K`" ;;
    language) LANGVAL="$VALUE" ;;
    spdx_id) LICENSEVAL="$VALUE" ;;
    date) DATEVAL="$VALUE" ;;
    message) MESSAGEVAL="`echo -e \"$VALUE\" | fold -s -w $WRAP | pr -o 1 -T `" ;;
   esac
  done
  IFS="$oldIFS" # Take '"' off IFS
  echo "______________________________________________________________________"
else
  fullpath="$repo/$path" # Remove deepest directory from path
  [ "${fullpath%/*}" == "$repo" ] && urlpath= || urlpath="${path%/*}"
  echo -e "!GophHub - $repo/$path\n______________________________________________________________________"
  echo -e "1..\t${REQUEST}?repo=${repo}${carry}&path=$urlpath"
fi

# List directory contents
#  Lists values into arrays before outputting directory list
oldIFS="$IFS"
IFS=\" # remove quotes from data using IFS
dirno=0
# Read in lines of JSON data with useful fields and parse them by splitting at the '"' character.
# Data for each directory entry is read into an array. The array index increments whenever an existing entry is found.
grep '"type":'$'\n''"name":'$'\n''"size":'$'\n''"path":'$'\n''"download_url":' git_dir.json \
| while read -r start FIELD sep VALUE comma || {
        # This part runs after the end of the end of output from Grep
        while [ $dirno -ge 0 ]
        do
         case "${TYPEVAL[$dirno]}" in
          file)  if [ $html -gt 0 -a ${SIZEVAL[$dirno]} -le $MAXSIZE ]
                then
                 # API HTML file download links
                 echo -e "h${NAMEVAL[$dirno]}\t$CGIPATH?repo=${repo}${carry}&file=1&path=${PATHVAL[$dirno]}"
                 elif [ $raw -gt 0 -a ${SIZEVAL[$dirno]} -le $MAXSIZE ]
                 then
                  # API raw file download links
                  echo -e "0${NAMEVAL[$dirno]}\t$CGIPATH?repo=${repo}${carry}&file=1&path=${PATHVAL[$dirno]}"
                 else
                  # Default to HTTPS file download links to raw.githubusercontent.com
                  echo -e "h${NAMEVAL[$dirno]}\tURL:${URLVAL[$dirno]}"
                 fi
              ;;
               # Directory links
          dir) echo -e "1${NAMEVAL[$dirno]}\t${REQUEST}?repo=${repo}${carry}&path=${PATHVAL[$dirno]}" ;;
         esac
         dirno=$[ $dirno - 1 ]
        done
        false
       }
 do
  case "$FIELD" in
   type) [ "${TYPEVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; TYPEVAL[$dirno]="$VALUE" ;;
   name) [ "${NAMEVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; NAMEVAL[$dirno]="$VALUE" ;;
   size) [ "${SIZEVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; sep="${sep#: *}"; SIZEVAL[$dirno]="${sep%,}" ;;
   path) [ "${PATHVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; PATHVAL[$dirno]="$VALUE" ;;
   download_url) [ "${URLVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; URLVAL[$dirno]="$VALUE" ;;
  esac
 done
IFS="$oldIFS" # Take '"' off IFS

# Display README, latest release info, and latest release download links
if [ -z "$path" ]
then
  echo '______________________________________________________________________'
  # Display default readme for the repo, with text wrapped and indented one space from the left.
  # Outputs one 'page' of $READMEMAX lines.
  [ $READMEMAX -gt 0 ] && fold -w $WRAP -s readme | pr -o 1 -T -l $READMEMAX +1:1
  echo '______________________________________________________________________'
  echo -e "hFull README in HTML\t$CGIPATH?repo=${repo}${carry}&file=readme"
  # Show latest release, if there is one
  if [ -f latest ]
  then
    echo '______________________________________________________________________'
    IFS=\" # remove quotes from data using IFS
    dirno=0
    # Read in lines of JSON data with useful fields and parse them by splitting at the '"' character
    # Size and browser_download_url values for each release file are held in an array.
    # Other fields should only appear once.
    grep '"browser_download_url":'$'\n''"size":'$'\n''"tag_name":'$'\n''"published_at":'$'\n''"tarball_url":'$'\n''"zipball_url":' latest \
    | while read -r start FIELD sep VALUE comma || {
        # This part runs after the end of the end of output from Grep
        echo -e "  Latest Release: $TAGVAL\n  Date: $DATEVAL\n  Release Files:"
        while [ $dirno -gt -1 ]
        do
          [ "$URLVAL" ] && echo -e "h${URLVAL[$dirno]##*/} (${SIZEVAL[$dirno]})\tURL:${URLVAL[$dirno]}"
          dirno=$[ $dirno - 1 ]
        done
        echo -e "hSource code TAR.GZ\tURL:$TARVAL\nhSource code ZIP\tURL:$ZIPVAL"
        false
       }
    do
      case "$FIELD" in
        tag_name) [ -z "$TAGVAL" ] && TAGVAL="$VALUE" ;;
        published_at) [ -z "$DATEVAL" ] && DATEVAL="$VALUE" ;;
        size) [ "${SIZEVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; sep="${sep#: *}"; SIZEVAL[$dirno]="`numfmt --to=iec \"${sep%,}\"`" ;;
        browser_download_url) [ "${URLVAL[$dirno]}" ] && dirno=$[ $dirno + 1 ]; URLVAL[$dirno]="$VALUE" ;;
        tarball_url) [ -z "$TARVAL" ] && TARVAL="$VALUE" ;;
        zipball_url) [ -z "$ZIPVAL" ] && ZIPVAL="$VALUE" ;;
      esac
    done
    IFS="$oldIFS" # Take '"' off IFS
  fi
  echo '______________________________________________________________________'
else
  echo -e '______________________________________________________________________\n'\
"1Repo Home\t${REQUEST}?repo=${repo}${carry}"
fi
echo -e "1Settings\t${REQUEST}?repo=${repo}${carry}&settings=1&path=${path}\n"\
"1GophHub Home\t$REQUEST?${carry#&}\n\n"\
"New files/directories accessed in the last $CACHETIME minutes:\n"\
"  $cachecount"

GophHub - $repo README

GophHub - ${fullpath%/*}/${path##*/}

GophHub - ${fullpath%/*}/${path##*/}

README empty or not found

File empty or not found

File empty or not found

GophHub - ${fullpath%/}/${path##/}

GophHub - ${fullpath%/}/${path##/}