From 4b3c7a6fa17f3e1b76e6f388b432c5483ae222ef Mon Sep 17 00:00:00 2001
From: James Taylor <jamest@uk.ibm.com>
Date: Thu, 11 Jan 2024 12:03:37 +0000
Subject: [PATCH] Add cicsdev-summary.sh script

Create script to generate summaries of the cicsdev stats

Signed-off-by: James Taylor <jamest@uk.ibm.com>
---
 README.md                  | 13 ++++++++-
 scripts/cicsdev-summary.sh | 55 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 67 insertions(+), 1 deletion(-)
 create mode 100755 scripts/cicsdev-summary.sh

diff --git a/README.md b/README.md
index 04855a70c4..30d4889676 100644
--- a/README.md
+++ b/README.md
@@ -1,7 +1,18 @@
-# repo-stats
+# CICSdev stats
 
 Download statistics for repos in the CICSDev org
 
+## Summary
+
+The [github-repo-stats GitHub Action](https://github.com/marketplace/actions/github-repo-stats) collects stats and generates [reports for each repository in the cicsdev organisation](https://github.com/cicsdev/repo-stats/tree/github-repo-stats/cicsdev) individually.
+
+The `cicsdev-summary.sh` script can be used to create summary files for *clones*, *forks*, *stargazers*, and *views*. To run the script:
+
+1. copy `cicsdev-summary.sh` to a suitable directory on your path, e.g. `~/bin`
+2. clone the `repo-stats` repository
+3. check out the `github-repo-stats` branch
+4. run `cicsdev-summary.sh` from the root of the `repo-stats` repository 
+
 ## TODO
 
 Get the list of repos automatically instead of hard coding it in the workflow
diff --git a/scripts/cicsdev-summary.sh b/scripts/cicsdev-summary.sh
new file mode 100755
index 0000000000..025b6bb8a8
--- /dev/null
+++ b/scripts/cicsdev-summary.sh
@@ -0,0 +1,55 @@
+#!/bin/sh
+
+statsdir=./cicsdev
+outdir=./summary
+tempdir="$outdir/temp"
+
+if ! command -v qsv >/dev/null 2>&1
+then
+    echo >&2 "This script requires qsv."
+    exit 1
+fi
+
+if [ ! -d "$statsdir" ]; then
+    echo >&2 "Cannot find cicsdev directory: $statsdir"
+    exit 1
+fi
+
+mkdir -p "$tempdir"
+
+for csvfile in forks stargazers; do
+    echo "Combining $csvfile.csv files..."
+    echo "repo,total" > "$tempdir/$csvfile.csv"
+    find "$statsdir" -name "$csvfile.csv" -type f -exec sh -c 'repo=$(echo "$1" | cut -d "/" -f3); printf "$repo,"; tail -n 1 $1 | cut -d , -f 2' _ {} \; >> "$tempdir/$csvfile.csv"
+    qsv sort -s repo "$tempdir/$csvfile.csv" > "$outdir/$csvfile.csv"
+    qsv sort -NR -s total "$tempdir/$csvfile.csv" | head -n 11 > "$outdir/${csvfile}_top10.csv"
+done
+
+for col in clones views; do
+    echo "Combining $col data..."
+    mkdir -p "$tempdir/$col"
+
+    find "$statsdir" -name "views_clones_aggregate.csv" -type f -exec sh -c 'repo=$(echo "$1" | cut -d "/" -f3); qsv select time_iso8601,$2_total "$1" | qsv rename date,$repo - > "$3/$repo.csv"' _ {} "$col" "$tempdir/$col" \;
+
+    rm -f "$tempdir/$col.csv"
+
+    for filename in "$tempdir/$col/"*.csv
+    do
+        if [ ! -f "$filename" ]; then
+            echo "Skipping $filename"
+            continue
+        elif [ ! -f "$tempdir/$col.csv" ]; then
+            echo "Copying $filename"
+
+            cp "$filename" "$tempdir/$col.csv"
+        else
+            echo "Joining $filename"
+
+            qsv joinp --null-value "0" --full date "$tempdir/$col.csv" date "$filename" > "$tempdir/joined_$col.csv"
+            mv "$tempdir/joined_$col.csv" "$tempdir/$col.csv"
+        fi
+    done
+
+    qsv sort -s date "$tempdir/$col.csv" > "$outdir/$col.csv"
+    qsv stats "$tempdir/$col.csv" | qsv select field,sum | qsv sort -NR -s sum | qsv rename repo,$col | head -n 11 > "$outdir/${col}_top10.csv"
+done