From 60a4a260f7952a768e06788a1e80468d90bc4cab Mon Sep 17 00:00:00 2001
From: paulzierep <paul.zierep@googlemail.com>
Date: Thu, 13 Jun 2024 12:52:06 +0200
Subject: [PATCH] Fix tool filtering (#118)

* fix wrong json merging

* allow empty or no tools_status file
---
 .github/workflows/fetch_all_tools.yaml |  4 ++--
 bin/extract_galaxy_tools.py            | 16 ++++++++++++++--
 bin/get_community_tools.sh             |  2 ++
 3 files changed, 18 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/fetch_all_tools.yaml b/.github/workflows/fetch_all_tools.yaml
index 218b3dfc..8e8ba610 100644
--- a/.github/workflows/fetch_all_tools.yaml
+++ b/.github/workflows/fetch_all_tools.yaml
@@ -77,9 +77,9 @@ jobs:
           python -m pip install -r requirements.txt
           sudo apt-get install jq
       - name: Merge all tools
-        run: | #merge files with only one header -> https://stackoverflow.com/questions/16890582/unixmerge-multiple-csv-files-with-same-header-by-keeping-the-header-of-the-firs
+        run: | #merge files with only one header -> https://stackoverflow.com/questions/16890582/unixmerge-multiple-csv-files-with-same-header-by-keeping-the-header-of-the-firs; map(.[]) -> https://stackoverflow.com/questions/42011086/merge-arrays-of-json (get flat array, one tool per entry)
           awk 'FNR==1 && NR!=1{next;}{print}' results/repositories*.list_tools.tsv > results/all_tools.tsv
-          jq -s '.' results/repositories*.list_tools.json > results/all_tools.json
+          jq -s 'map(.[])' results/repositories*.list_tools.json > results/all_tools.json 
       - name: Wordcloud and interactive table
         run: |
           bash ./bin/extract_all_tools_downstream.sh
diff --git a/bin/extract_galaxy_tools.py b/bin/extract_galaxy_tools.py
index aa71d70a..631e5f9d 100644
--- a/bin/extract_galaxy_tools.py
+++ b/bin/extract_galaxy_tools.py
@@ -769,8 +769,20 @@ def reduce_ontology_terms(terms: List, ontology: Any) -> List:
             tools = json.load(f)
         # get categories and tools to exclude
         categories = read_file(args.categories)
-        status = pd.read_csv(args.status, sep="\t", index_col=0, header=None).to_dict("index")
+        try:
+            status = pd.read_csv(args.status, sep="\t", index_col=0, header=None).to_dict("index")
+        except Exception as ex:
+            print(f"Failed to load tool_status.tsv file with:\n{ex}")
+            print("Not assigning tool status for this community !")
+            status = {}
+
         # filter tool lists
         ts_filtered_tools, filtered_tools = filter_tools(tools, categories, status)
+
         export_tools_to_tsv(ts_filtered_tools, args.ts_filtered_tools, format_list_col=True)
-        export_tools_to_tsv(filtered_tools, args.filtered_tools, format_list_col=True)
+
+        # if there are no filtered tools return the ts filtered tools
+        if filtered_tools:
+            export_tools_to_tsv(filtered_tools, args.filtered_tools, format_list_col=True)
+        else:
+            export_tools_to_tsv(ts_filtered_tools, args.filtered_tools, format_list_col=True)
diff --git a/bin/get_community_tools.sh b/bin/get_community_tools.sh
index a0566eef..efb03f9c 100755
--- a/bin/get_community_tools.sh
+++ b/bin/get_community_tools.sh
@@ -1,5 +1,7 @@
 #!/usr/bin/env bash
 
+# stop on error
+set -e 
 
 for com_data_fp in data/communities/* ; do
         if [[ -d "$com_data_fp" && ! -L "$com_data_fp" ]]; then