Johnpreed/do not merge

README.md

@@ -14,3 +14,4 @@ Make a pull request and we'll consider it.
 * _graphql_: here's a bunch of sample GraphQL queries that can be run against our [GitHub GraphQL API](https://developer.github.com/early-access/graphql).
 * _hooks_: want to find out how to write a consumer for [our web hooks](https://developer.github.com/webhooks/)? The examples in this subdirectory show you how. We are open for more contributions via pull requests.
 * _pre-receive-hooks_: this one contains [pre-receive-hooks](https://help.github.com/enterprise/admin/guides/developer-workflow/about-pre-receive-hooks/) that can block commits on GitHub Enterprise that do not fit your requirements. Do you have more great examples? Create a pull request and we will check it out.
+* _scripts_: want to analyze or clean-up your Git repository? The scripts in this subdirectory show you how. We are open for more contributions via pull requests.

scripts/README.md

@@ -0,0 +1,17 @@
+# Git Repo Analysis Scripts
+
+Git can become slow if a repository exceeds certain thresholds ([read this for details](http://larsxschneider.github.io/2016/09/21/large-git-repos)). Use the scripts explained below to identify possible culprits in a repository. The scripts have been tested on macOS but they should run on Linux as is.
+
+_Hint:_ The scripts can run for a long time and output a lot lines. Pipe their output to a file (`./script > myfile`) for further processing.
+
+## Large by File Size
+Use the [git-find-large-files](git-find-large-files) script to identity large files in your Git repository that you could move to [Git LFS](https://git-lfs.github.com/) (e.g. using [git-lfs-migrate](https://github.com/git-lfs/git-lfs/blob/master/docs/man/git-lfs-migrate.1.ronn)).
+
+Use the [git-find-lfs-extensions](git-find-lfs-extensions) script to identify certain file types that you could move to [Git LFS](https://git-lfs.github.com/).
+
+## Large by File Count
+Use the [git-find-dirs-many-files](git-find-dirs-many-files) and [git-find-dirs-unwanted](git-find-dirs-unwanted) scripts to identify directories with a large number of files. These might indicate 3rd party components that could be extracted.
+
+Use the [git-find-dirs-deleted-files](git-find-dirs-deleted-files) to identify directories that have been deleted and used to contain a lot of files. If you purge all files under these directories from your history then you might be able significantly reduce the overall size of your repository.
+
+

scripts/git-change-author

@@ -0,0 +1,29 @@
+#!/usr/bin/env bash
+#
+# Fix an invalid committer/author all commits of your repository.
+#
+# Usage:
+# git-change-author <old-email> <new-name> <new-email>
+#
+# Author: Lars Schneider, https://github.com/larsxschneider
+#
+
+filter=$(cat <<EOF
+    OLD_EMAIL='$1'
+    NEW_NAME='$2'
+    NEW_EMAIL='$3'
+
+    if [ "\$GIT_COMMITTER_EMAIL" = "\$OLD_EMAIL" ]
+    then
+        export GIT_COMMITTER_NAME="\$NEW_NAME"
+        export GIT_COMMITTER_EMAIL="\$NEW_EMAIL"
+    fi
+    if [ "\$GIT_AUTHOR_EMAIL" = "\$OLD_EMAIL" ]
+    then
+        export GIT_AUTHOR_NAME="\$NEW_NAME"
+        export GIT_AUTHOR_EMAIL="\$NEW_EMAIL"
+    fi
+EOF
+)
+
+git filter-branch --env-filter "$filter" --tag-name-filter cat -- --all

scripts/git-find-dirs-deleted-files

@@ -0,0 +1,34 @@
+#!/usr/bin/env bash
+#
+# Print the number of deleted files per directory. The output indicates
+# if the directory is present in the HEAD revision.
+#
+# A deleted directory with a lot of files could indicate a 3rd party
+# component that has been deleted. These are usually good candidates for
+# purging to make Git repositories smaller (see `git-purge-files`).
+#
+# The script must be called from the root of the Git repository.
+#
+# Usage:
+# git-find-dirs-deleted-files
+#
+# Output: [deleted file count] [directory still in HEAD revision?] [directory]
+#
+# Author: Lars Schneider, https://github.com/larsxschneider
+#
+
+git -c diff.renameLimit=30000 log --diff-filter=D --summary |
+    grep ' delete mode ...... ' |
+    sed 's/ delete mode ...... //' |
+    while read -r F ; do
+        D=$(dirname "$F");
+        if ! [ -d "$D" ]; then
+            while ! [ -d "$(dirname "$D")" ] ; do D=$(dirname "$D"); done;
+            echo "deleted $D";
+        else
+            echo "present $D";
+        fi;
+    done |
+    sort |
+    uniq -c |
+    sort -k 2,2 -r

scripts/git-find-dirs-many-files

@@ -0,0 +1,31 @@
+#!/usr/bin/env bash
+#
+# Print directories with the number of files underneath them.
+#
+# A directory with a lot of files could indicates a 3rd party component.
+# These are usually good candidates for purging to make Git repositories
+# smaller (see `git-purge-files`).
+#
+# The script must be called from the root of the Git repository.
+#
+# Usage:
+# git-find-dirs-many-files [file count threshold]
+#
+# Author: Lars Schneider, https://github.com/larsxschneider
+#
+
+if [ -z "$1" ]; then
+    FILE_COUNT=100
+else
+    FILE_COUNT=$1
+fi
+
+IFS=$'\n';
+DIRS=$(find . -type d -not -path "./.git/*" -exec bash -c 'COUNT=$(find "$0" -type f | wc -l); echo "$COUNT $0"' {} \; | sort -nr)
+
+for DIR in $DIRS; do
+    if [ $(($(echo $DIR | sed 's/\..*//'))) -le $FILE_COUNT ]; then
+        break
+    fi
+    echo $DIR
+done

scripts/git-find-dirs-unwanted

@@ -0,0 +1,65 @@
+#!/usr/bin/env bash
+#
+# git-find-dirs-unwanted.sh
+#
+# Search the entire history of a Git repository for (potentially)
+# unwanted directories. E.g. 3rd party directories, temp, build or
+# Perforce stream directories.
+#
+# The script prints the number of files under each directory to see the
+# impact on the Git tree. Directories with a large number of files can
+# be good candidates for exclusions in repository migrations to Git.
+#
+# The script must be called in the Git root directory.
+#
+# Author: Lars Schneider, https://github.com/larsxschneider
+#
+
+DIRS=$(git -c diff.renameLimit=30000 log --all --name-only --pretty=format: \
+    | awk -F'[^/]*$' '{print $1}' \
+    | sort -u \
+    | grep -i \
+        -e 3p \
+        -e 3rd \
+        -e artifacts \
+        -e assemblies \
+        -e backup \
+        -e bin \
+        -e build \
+        -e components \
+        -e debug \
+        -e deploy \
+        -e generated \
+        -e install \
+        -e lib \
+        -e modules \
+        -e obj \
+        -e output \
+        -e packages \
+        -e party \
+        -e recycle.bin \
+        -e release \
+        -e resources \
+        -e streams \
+        -e temp \
+        -e third \
+        -e tmp \
+        -e tools \
+        -e util \
+        -e vendor \
+        -e x64 \
+        -e x86 \
+)
+
+IFS=$'\n'
+for I in $DIRS; do
+    if [ -e "$I" ]; then
+        FILE_COUNT=$(find "$I" -type f | wc -l)
+        echo "$FILE_COUNT $I"
+    else
+        while ! [ -e $(dirname "$I") ]; do
+            I=$(dirname "$I")/;
+        done;
+        echo "deleted  $I"
+    fi
+done | sort -n -r | uniq

scripts/git-find-ignored-files

@@ -0,0 +1,63 @@
+#!/usr/bin/env bash
+#
+# Find all files present in the index and working tree ignored by .gitignore.
+#
+# Usage:
+# git-find-ignored-files [-s | --sort-by-size] [--help]
+#
+# Author: Patrick Lühne, https://www.luehne.de/
+#
+
+function print_help
+{
+	grep "^# Usage" < "$0" | cut -c 3-
+}
+
+if [[ $# -gt 1 ]]
+then
+	print_help
+	exit 1
+fi
+
+case "$1" in
+	-h|--help)
+		print_help
+		exit 0
+		;;
+	-s|--sort-by-size)
+		;;
+	*)
+		if [[ $# -gt 0 ]]
+		then
+			(>&2 echo "error: unknown option “$1”")
+			print_help
+			exit 1
+		fi
+		;;
+esac
+
+# Find all ignored files
+files=$(git ls-files --ignored --exclude-standard)
+
+# Stop if no ignored files were found
+if [[ -z $files ]]
+then
+	(>&2 echo "info: no ignored files in working tree or index")
+	exit 0
+fi
+
+# Compute the file sizes of all these files
+file_sizes=$(echo "$files" | tr '\n' '\0' | xargs -0 du -sh)
+
+# Obtain the origins why these files are ignored
+gitignore_origins=$(echo "$files" | git check-ignore --verbose --stdin --no-index)
+
+# Merge the two lists into one
+command="join -1 2 -2 2 -t $'\t' -o 1.1,1.2,2.1 <(echo \"$file_sizes\") <(echo \"$gitignore_origins\")"
+
+if [[ $1 =~ ^-s|--sort-by-size$ ]]
+then
+	command="$command | sort -h"
+fi
+
+eval "$command"

scripts/git-find-large-files

@@ -0,0 +1,91 @@
+#!/usr/bin/env bash
+#
+# Print the largest files in a Git repository. The script must be called
+# from the root of the Git repository. You can pass a threshold to print
+# only files greater than a certain size (compressed size in Git database,
+# default is 500kb).
+#
+# Files that have a large compressed size should usually be stored in
+# Git LFS [2].
+#
+# Based on script from Antony Stubbs [1] and improved with ideas from Peff.
+#
+# [1] http://stubbisms.wordpress.com/2009/07/10/git-script-to-show-largest-pack-objects-and-trim-your-waist-line/
+# [2] https://git-lfs.github.com/
+#
+# Usage:
+# git-find-large-files [size threshold in KB]
+#
+# Author: Lars Schneider, https://github.com/larsxschneider
+#
+
+if [ -z "$1" ]; then
+    MIN_SIZE_IN_KB=500
+else
+    MIN_SIZE_IN_KB=$1
+fi
+
+# Use "look" if it is available, otherwise use "grep" (e.g. on Windows)
+if look >/dev/null 2>&1; then
+    # On Debian the "-b" is available and required to make "look" perform
+    # a binary search (see https://unix.stackexchange.com/a/499312/275508 ).
+    if look 2>&1 | grep -q .-b; then
+        search="look -b"
+    else
+        search=look
+    fi
+else
+    search=grep
+fi
+
+# set the internal field separator to line break,
+# so that we can iterate easily over the verify-pack output
+IFS=$'\n';
+
+# list all objects including their size, sort by compressed size
+OBJECTS=$(
+    git cat-file \
+        --batch-all-objects \
+        --batch-check='%(objectsize:disk) %(objectname)' \
+    | sort -nr
+)
+
+TMP_DIR=$(mktemp -d "${TMPDIR:-/tmp}/git-find-large-files.XXXXXX") || exit
+trap "rm -rf '$TMP_DIR'" EXIT
+
+git rev-list --all --objects | sort > "$TMP_DIR/objects"
+git rev-list --all --objects --max-count=1 | sort > "$TMP_DIR/objects.1"
+
+for OBJ in $OBJECTS; do
+    # extract the compressed size in kilobytes
+    COMPRESSED_SIZE=$(($(echo $OBJ | cut -f 1 -d ' ')/1024))
+
+    if [ $COMPRESSED_SIZE -le $MIN_SIZE_IN_KB ]; then
+        break
+    fi
+
+    # extract the SHA
+    SHA=$(echo $OBJ | cut -f 2 -d ' ')
+
+    # find the objects location in the repository tree
+    LOCATION=$($search $SHA "$TMP_DIR/objects" | sed "s/$SHA //")
+    if $search $SHA "$TMP_DIR/objects.1" >/dev/null; then
+        # Object is in the head revision
+        HEAD="Present"
+    elif [ -e $LOCATION ]; then
+        # Objects path is in the head revision
+        HEAD="Changed"
+    else
+        # Object nor its path is in the head revision
+        HEAD="Deleted"
+    fi
+
+    echo "$COMPRESSED_SIZE,$HEAD,$LOCATION" >> "$TMP_DIR/output"
+done
+
+if [ -f "$TMP_DIR/output" ]; then
+    column -t -s ',' < "$TMP_DIR/output"
+fi
+
+rm -rf "$TMP_DIR"
+exit 0