Merge pull request #174 from tomice/master

* commitsByYear() previously could only support Default-style
  dates for git log. However, a user can use a variety of different
  date formats such as relative, default, iso, rfc, short, and raw.

  This change modifies commitsByYear() to handle multiple different
  date strings

Fixes #172

README.md

@@ -6,7 +6,7 @@
 >
 > Any git repository may contain tons of information about commits, contributors, and files. Extracting this information is not always trivial, mostly because there are a gadzillion options to a gadzillion git commands - I don't think there is a single person alive who knows them all. Probably not even [Linus Torvalds](https://github.com/torvalds) himself :).
 
-![mainMenuScreenshot](https://user-images.githubusercontent.com/8818630/154823670-f42a111b-45f4-47fc-aea7-80059827c8e6.png)
+![mainMenuScreenshot](https://github.com/user-attachments/assets/05a20ff1-44f6-4e44-9d62-c2c089f9ff4e)
 
 ## Table of Contents
 
@@ -54,9 +54,9 @@
 
 ## Screenshots
 
-![commitsByWeekdayScreenshot](https://user-images.githubusercontent.com/8818630/154823677-e34ca867-ae61-4755-bf89-c877f809c591.png)
+![commitsByWeekdayScreenshot](https://github.com/user-attachments/assets/3a55f3ac-8801-4bbf-9b3a-92b53a64631e)
 
-![commitsByHourScreenshot](https://user-images.githubusercontent.com/8818630/154823679-bdadc26c-c644-4a01-b625-85e330f85d41.png)
+![commitsByHourScreenshot](https://github.com/user-attachments/assets/d7de5280-8bb9-4391-9c6c-7e688f2df171)
 
 ## Usage
 
@@ -116,6 +116,8 @@ LIST OPTIONS
         show branches by date
     -C, --contributors
         see a list of everyone who contributed to the repo
+    -n, --new-contributors
+        list everyone who made their first contribution since a specified date
     -a, --commits-per-author
         displays a list of commits per author
     -d, --commits-per-day
@@ -210,7 +212,7 @@ You can change to the legacy color scheme by toggling the variable `_MENU_THEME`
 export _MENU_THEME="legacy"
 ```
 
-![legacyThemeScreenshot](https://user-images.githubusercontent.com/8818630/154823711-3dd0c268-f3cb-42e5-9094-0eb8e45d1761.png)
+![legacyThemeScreenshot](https://github.com/user-attachments/assets/3b319c1a-827f-47b8-bbfa-b8b59a39deef)
 
 ## Installation
 

git-quick-stats

@@ -17,7 +17,7 @@ _since=${_GIT_SINCE:-}
 if [[ -n "${_since}" ]]; then
     _since="--since=$_since"
 else
-    _since="--since=$(git log --reverse --format='%ad' | head -n1)"
+    _since="--since=$(git log --reverse --format='%ad' --date=iso | head -n1)"
 fi
 
 # End of git log date. Respects all git datetime formats
@@ -182,6 +182,8 @@ LIST OPTIONS
         show branches by date
     -C, --contributors
         see a list of everyone who contributed to the repo
+    -n, --new-contributors
+        list everyone who made their first contribution since a specified date
     -a, --commits-per-author
         displays a list of commits per author
     -d, --commits-per-day
@@ -274,18 +276,19 @@ function showMenu() {
     printf %b "${NUMS}    8)${TEXT} Branch tree view (last $_limit)\\n"
     printf %b "${NUMS}    9)${TEXT} All branches (sorted by most recent commit)\\n"
     printf %b "${NUMS}   10)${TEXT} All contributors (sorted by name)\\n"
-    printf %b "${NUMS}   11)${TEXT} Git commits per author\\n"
-    printf %b "${NUMS}   12)${TEXT} Git commits per date\\n"
-    printf %b "${NUMS}   13)${TEXT} Git commits per month\\n"
-    printf %b "${NUMS}   14)${TEXT} Git commits per year\\n"
-    printf %b "${NUMS}   15)${TEXT} Git commits per weekday\\n"
-    printf %b "${NUMS}   16)${TEXT} Git commits per weekday by author\\n"
-    printf %b "${NUMS}   17)${TEXT} Git commits per hour\\n"
-    printf %b "${NUMS}   18)${TEXT} Git commits per hour by author\\n"
-    printf %b "${NUMS}   19)${TEXT} Git commits per timezone\\n"
-    printf %b "${NUMS}   20)${TEXT} Git commits per timezone by author\\n"
+    printf %b "${NUMS}   11)${TEXT} New contributors (sorted by email)\\n"
+    printf %b "${NUMS}   12)${TEXT} Git commits per author\\n"
+    printf %b "${NUMS}   13)${TEXT} Git commits per date\\n"
+    printf %b "${NUMS}   14)${TEXT} Git commits per month\\n"
+    printf %b "${NUMS}   15)${TEXT} Git commits per year\\n"
+    printf %b "${NUMS}   16)${TEXT} Git commits per weekday\\n"
+    printf %b "${NUMS}   17)${TEXT} Git commits per weekday by author\\n"
+    printf %b "${NUMS}   18)${TEXT} Git commits per hour\\n"
+    printf %b "${NUMS}   19)${TEXT} Git commits per hour by author\\n"
+    printf %b "${NUMS}   20)${TEXT} Git commits per timezone\\n"
+    printf %b "${NUMS}   21)${TEXT} Git commits per timezone by author\\n"
     printf %b "\\n${TITLES} Suggest:\\n"
-    printf %b "${NUMS}   21)${TEXT} Code reviewers (based on git history)\\n"
+    printf %b "${NUMS}   22)${TEXT} Code reviewers (based on git history)\\n"
     printf %b "\\n${HELP_TXT}Please enter a menu option or ${EXIT_TXT}press Enter to exit.\\n"
     printf %b "${TEXT}> ${NORMAL}"
     read -r opt
@@ -337,11 +340,15 @@ function detailedGitStats() {
         if(more["total"] > 0) {
             printf "\t  insertions:    %d\t(%.0f%%)\n", more[author], \
                 (more[author] / more["total"] * 100)
+        } else {
+            printf "\t  insertions:    %d\t(%.0f%%)\n", 0, 0
         }
 
         if(less["total"] > 0) {
             printf "\t  deletions:     %d\t(%.0f%%)\n", less[author], \
                 (less[author] / less["total"] * 100)
+        } else {
+            printf "\t  deletions:     %d\t(%.0f%%)\n", 0, 0
         }
 
         if(file["total"] > 0) {
@@ -498,11 +505,15 @@ function csvOutput() {
         if(more["total"] > 0) {
             printf "%d,%.0f%%,", more[author], \
                 (more[author] / more["total"] * 100)
+        } else {
+            printf "0,0%%,"
         }
 
         if(less["total"] > 0) {
             printf "%d,%.0f%%,", less[author], \
                 (less[author] / less["total"] * 100)
+        } else {
+            printf "0,0%%,"
         }
 
         if(file["total"] > 0) {
@@ -685,6 +696,26 @@ function contributors() {
         --format='%aN' $_log_options $_pathspec | sort -u | cat -n
 }
 
+################################################################################
+# DESC: Lists all new contributors to a repo since the specified time
+# ARGS: $newDate (required): Cutoff date for being considered "new"
+# OUTS: None
+################################################################################
+function newContributors() {
+    optionPicked "New contributors since $newDate:"
+    local contributors=$(git -c log.showSignature=false log --use-mailmap $_merges \
+                             "$_since" "$_until" --format='%aE' $_log_options \
+                             $_pathspec | sort -u)
+    for c in $contributors; do
+        local firstCommit=$(git -c log.showSignature=false log --author="$c" \
+                         --reverse --use-mailmap $_merges "$_since" "$_until" \
+                         --format='%at' $_log_options $_pathspec | head -n 1)
+        if [[ $firstCommit -ge $(date -d "$newDate" +%s) ]]; then
+           echo "$c"
+        fi
+    done
+}
+
 ################################################################################
 # DESC: Displays the number of commits and percentage contributed to the repo
 #       per author and sorts them by contribution percentage
@@ -730,6 +761,47 @@ function commitsPerDay() {
         --date=short --format='%ad' $_log_options $_pathspec | sort | uniq -c
 }
 
+################################################################################
+# DESC: Convert a timestamp to a date string to handle git's date formats
+# ARGS: $1: Timestamp
+# OUTS: Echoes a four-digit year
+################################################################################
+function parse_year() {
+    local date_str="$1"
+    local year
+    local timestamp
+    local default_git_date_regex
+
+    # Handle the raw UNIX timestamp format i.e. 1697375696 +0000
+    if [[ "$date_str" =~ ^[0-9]+(\ [+-][0-9]{4})?$ ]]; then
+        timestamp=$(echo "$date_str" | awk '{print $1}')
+        year=$(date -d "@$timestamp" '+%Y' 2>/dev/null)
+    else
+        # Default case can get funky. We need to create a clever regex to
+        # handle the default case which is like Mon Oct 15 12:34:56 2023 +0000
+        # Let's make this explicit for future devs to follow along.
+        default_git_date_regex='^' # Start from the beginning of the string
+        default_git_date_regex+='[A-Za-z]{3}\ ' # Day abbrev
+        default_git_date_regex+='[A-Za-z]{3}\ ' # Month abbrev
+        default_git_date_regex+='[0-9]{1,2}\ ' # Day of the month
+        default_git_date_regex+='[0-9]{2}:[0-9]{2}:[0-9]{2}\ ' # Time HH:MM:SS
+        default_git_date_regex+='[0-9]{4}\ ' # Year
+        default_git_date_regex+='[+-][0-9]{4}$' # Timezone offset
+
+        if [[ "$date_str" =~ $default_git_date_regex ]]; then
+            # Move the year before the time to match a format that Date can parse
+            date_str=$(echo "$date_str" | awk '{print $1, $2, $3, $5, $4, $6}')
+        elif [[ "$date_str" =~ ^[0-9]{1,2}/[0-9]{1,2}/[0-9]{2,4} ]]; then
+            # Handle DD/MM/YYYY format
+            date_str=$(echo "$date_str" | awk -F'/' '{print $2"/"$1"/"$3}')
+        fi
+        # Extract the final date
+        year=$(date -d "$date_str" '+%Y' 2>/dev/null)
+    fi
+
+    echo "$year"
+}
+
 ################################################################################
 # DESC: Displays a horizontal bar graph based on total commits per year
 # ARGS: None
@@ -738,30 +810,38 @@ function commitsPerDay() {
 function commitsByYear() {
     optionPicked "Git commits by year:"
     local year startYear endYear __since __until
-    startYear=$(echo "$_since" | sed -E 's/^.* ([0-9]{4})( .*)?$/\1/')
-    endYear=$(echo "$_until" | sed -E 's/^.* ([0-9]{4})( .*)?$/\1/')
 
+    # Extract the date strings from $_since and $_until
+    since_date="${_since#--since=}"
+    until_date="${_until#--until=}"
+
+    # Grab the four digit year from $_since and $_until
+    startYear=$(parse_year "$since_date")
+    endYear=$(parse_year "$until_date")
+    
     echo -e "\tyear\tsum"
-    for year in $(seq "$startYear" "$endYear")
-    do
-        if [ "$year" = "$startYear" ]
-        then
-          __since=$_since
-          __until="--until=$year-12-31"
-        elif [ "$year" = "$endYear" ]
-        then
-          __since="--since=$year-01-01"
-          __until=$_until
+    # Add time strings to make these a touch more robust
+    for year in $(seq "$startYear" "$endYear"); do
+        if [[ "$year" = "$startYear" ]]; then
+            __since=$_since
+            __until="--until=$year-12-31 23:59:59"
+        elif [[ "$year" = "$endYear" ]]; then
+            __since="--since=$year-01-01 00:00:00"
+            __until=$_until
         else
-          __since="--since=$year-01-01"
-          __until="--until=$year-12-31"
+            __since="--since=$year-01-01 00:00:00"
+            __until="--until=$year-12-31 23:59:59"
         fi
-
-        echo -en "\t$year\t"
-        git -c log.showSignature=false shortlog -n $_merges --format='%ad %s' \
-            "$__since" "$__until" $_log_options | grep -cE \
-              " \w\w\w [0-9]{1,2} [0-9][0-9]:[0-9][0-9]:[0-9][0-9] $year " \
-                || continue
+        
+        # Count commits directly using git rev-list instead of git log
+        commit_count=$(
+            git rev-list --count $_merges \
+            "$__since" "$__until" HEAD $_log_options
+        )
+        echo -e "\t$year\t$commit_count"
+        # TODO: The bar graph can get funky when there are only a handful of
+        # commits. We can set a max length to try to fix this, but this is a
+        # bit of a problem across all the bar graphs.
     done | awk '{
         count[$1] = $2
         total += $2
@@ -955,7 +1035,11 @@ function suggestReviewers() {
 checkUtils
 
 # Check if we are currently in a git repo.
-git rev-parse --is-inside-work-tree > /dev/null
+if ! git rev-parse --is-inside-work-tree > /dev/null; then
+    echo "ERROR: You need to be inside a git repo to parse stats!"
+    usage
+    exit 1
+fi
 
 # Parse non-interative commands
 if [[ "$#" -eq 1 ]]; then
@@ -1004,6 +1088,20 @@ if [[ "$#" -eq 1 ]]; then
         -b|--branch-tree) branchTree;;
         -D|--branches-by-date) branchesByDate;;
         -C|--contributors) contributors;;
+        -n|--new-contributors)
+            newDate=""
+            while [[ -z "${newDate}" ]]; do
+                read -r -p "Since what date? (e.g. '2023-04-13', '13 April 2023', 'last Thursday') " newDate
+                # Test if the date provide is valid and try again if it isn't.
+                # date(1) is pretty accepting of time stamps but you never know
+                # what people may try. This script doesn't provide any additional
+                # output for a bad date since `date`'s STDERR already contains
+                # useful information.
+                if ! date -d "${newDate}" +%s > /dev/null 2>&1; then
+                   newDate=""
+                fi
+            done
+            newContributors "${newDate}";;
         -a|--commits-per-author) commitsPerAuthor;;
         -d|--commits-per-day) commitsPerDay;;
         -Y|--commits-by-year ) commitsByYear;;
@@ -1083,29 +1181,38 @@ while [[ "${opt}" != "" ]]; do
         8) branchTree; showMenu;;
         9) branchesByDate; showMenu;;
        10) contributors; showMenu;;
-       11) commitsPerAuthor; showMenu;;
-       12) commitsPerDay; showMenu;;
-       13) commitsByMonth; showMenu;;
-       14) commitsByYear; showMenu;;
-       15) commitsByWeekday; showMenu;;
-       16) author=""
+       11) newDate=""
+           while [[ -z "${newDate}" ]]; do
+               read -r -p "Since what date? (e.g. '2023-04-13', '13 April 2023', 'last Thursday') " newDate
+               # Test if the date provide is valid and try again if it isn't.
+               if ! date -d "${newDate}" +%s > /dev/null 2>&1; then
+                  newDate=""
+               fi
+           done
+           newContributors "${newDate}"; showMenu;;
+       12) commitsPerAuthor; showMenu;;
+       13) commitsPerDay; showMenu;;
+       14) commitsByMonth; showMenu;;
+       15) commitsByYear; showMenu;;
+       16) commitsByWeekday; showMenu;;
+       17) author=""
            while [[ -z "${author}" ]]; do
                read -r -p "Which author? " author
            done
            commitsByWeekday "${author}"; showMenu;;
-       17) commitsByHour; showMenu;;
-       18) author=""
+       18) commitsByHour; showMenu;;
+       19) author=""
            while [[ -z "${author}" ]]; do
                read -r -p "Which author? " author
            done
            commitsByHour "${author}"; showMenu;;
-       19) commitsByTimezone; showMenu;;
-       20) author=""
+       20) commitsByTimezone; showMenu;;
+       21) author=""
            while [[ -z "${author}" ]]; do
                read -r -p "Which author? " author
            done
            commitsByTimezone "${author}"; showMenu;;
-       21) suggestReviewers; showMenu;;
+       22) suggestReviewers; showMenu;;
        q|"\n") exit;;
        *) clear; optionPicked "Pick an option from the menu"; showMenu;;
     esac

git-quick-stats.1

@@ -70,6 +70,11 @@ show branches by date
 see a list of everyone who contributed to the repo
 .HP
 .PP
+\fB\-n\fR, \fB\-\-new\-contributors\fR
+.IP
+list everyone who made their first contribution since a specified date
+.HP
+.PP
 \fB\-a\fR, \fB\-\-commits\-per\-author\fR
 .IP
 displays a list of commits per author

tests/commands_test.sh

@@ -50,6 +50,8 @@ LIST OPTIONS
         show branches by date
     -C, --contributors
         see a list of everyone who contributed to the repo
+    -n, --new-contributors
+        list everyone who made their first contribution since a specified date
     -a, --commits-per-author
         displays a list of commits per author
     -d, --commits-per-day