Skip to content

Commit

Permalink
feat: Simplify Vespa documentation feed workflow
Browse files Browse the repository at this point in the history
  • Loading branch information
esolitos authored and gitbutler-client committed Nov 8, 2024
1 parent eff0f90 commit 5b24755
Showing 1 changed file with 14 additions and 80 deletions.
94 changes: 14 additions & 80 deletions .github/workflows/feed.yml
Original file line number Diff line number Diff line change
@@ -1,86 +1,20 @@
name: Vespa Documentation Search Feed

on:
push:
branches: [ master ]
paths:
- '*.md'
- '*.html'
- '**/*.md'
- '**/*.html'
- '.github/workflows/feed.yml'

env:
DATA_PLANE_PUBLIC_KEY : ${{ secrets.VESPA_TEAM_DATA_PLANE_PUBLIC_CERT }}
DATA_PLANE_PRIVATE_KEY : ${{ secrets.VESPA_TEAM_DATA_PLANE_PRIVATE_KEY }}
VESPA_CLI_DATA_PLANE_CERT : ${{ secrets.VESPA_TEAM_VESPA_CLI_DATA_PLANE_CERT }}
VESPA_CLI_DATA_PLANE_KEY : ${{ secrets.VESPA_TEAM_VESPA_CLI_DATA_PLANE_KEY }}
branches:
- main
- update-vespa-documentation-search-feed-workflow # Temporary branch for testing

jobs:
build:
runs-on: ubuntu-latest
steps:

- uses: actions/checkout@v4

- uses: ruby/setup-ruby@v1
with:
ruby-version: 3.1
bundler-cache: false

- name: Build site
run: |
bundle install
bundle exec jekyll build -p _plugins-vespafeed
- name: Setup Python
uses: actions/setup-python@v5
with:
python-version: '3.x'

- name: Install dependencies
run: |
pip3 install PyYAML mmh3 requests html5lib beautifulsoup4 markdownify tiktoken
- name: Install Vespa CLI
uses: vespa-engine/setup-vespa-cli-action@v1

- name: Feed docs site
run: |
# The python scripts below uses the Vespa CLI for feeding / data access.
# See https://docs.vespa.ai/en/vespa-cli.html.
# The environment variables below have credentials for endpoint access -
# use the key/cert files in .vespa and paste their content into GitHub Secrets.
# VESPA_CLI_API_KEY must be set and empty as below.
export VESPA_CLI_DATA_PLANE_CERT
export VESPA_CLI_DATA_PLANE_KEY
export VESPA_CLI_API_KEY=
./feed_to_vespa.py _config.yml
- name: Feed paragraphs site
run: |
export VESPA_CLI_DATA_PLANE_CERT
export VESPA_CLI_DATA_PLANE_KEY
export VESPA_CLI_API_KEY=
./feed-split.py open_index.json https://docs.vespa.ai questions.jsonl
./feed_to_vespa.py _paragraphs_config.yml
- name: Feed suggestions
run: |
export VESPA_CLI_DATA_PLANE_CERT
export VESPA_CLI_DATA_PLANE_KEY
export VESPA_CLI_API_KEY=
./feed_to_vespa.py _suggestions_config.yml

- name: Generate and feed reference suggestions
run: |
./generate_suggestions_from_reference.py en/reference/schema-reference.html https://docs.vespa.ai/ open-p > suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-admin.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-container.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-docproc.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-processing.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-search.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-http.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./generate_suggestions_from_reference.py en/reference/services-content.html https://docs.vespa.ai/ open-p >> suggestions.jsonl
./convert.py suggestions.jsonl suggestions_reference_index.json
./feed_to_vespa.py _suggestions_reference_config.yml
uses: vespa-engine/gh-actions/.github/workflows/jekyll-feed-to-vespa.yml@fix-jekyll-build
with:
hook-post-build-script: |
echo "Find all JSON files"
find . -type f -name '*.json'
echo "\n"
echo "Find the open_index.json file"
find . -type f -name '*open_index*'
ls -l /github/workspace/_site/json_indexes/open_index.json

0 comments on commit 5b24755

Please sign in to comment.