DEV-458 end-to-end automated test for metadata workflow

.gitignore

@@ -22,3 +22,5 @@ compare_*
 *.jsonl
 t/fixtures/rights_dbm
 cover_db
+**/.*.sw?
+*~

Dockerfile

@@ -21,6 +21,7 @@ RUN apt-get update && apt-get install -y \
     libmarc-perl \
     libmarc-record-perl \
     libmarc-xml-perl \
+    libmariadb-dev \
     libnet-ssleay-perl \
     libtest-output-perl \
     libwww-perl \
@@ -39,7 +40,7 @@ RUN cpanm --notest \
 # Ruby setup
 ENV BUNDLE_PATH /gems
 ENV RUBYLIB /usr/src/app/lib
-RUN gem install bundler
+RUN gem install bundler --version "~> 2.5.23"
 RUN bundle config --global silence_root_warning 1
 RUN bundle install
 

Gemfile

@@ -2,10 +2,18 @@
 
 source "https://rubygems.org"
 
+gem "canister"
+gem "dotenv"
+gem "faraday"
+gem "mysql2"
+gem "sequel"
+
 group :development, :test do
+  gem "climate_control"
   gem "pry"
   gem "rspec"
   gem "simplecov"
   gem "simplecov-lcov"
   gem "standardrb"
+  gem "webmock"
 end

Gemfile.lock

@@ -1,88 +1,124 @@
 GEM
   remote: https://rubygems.org/
   specs:
+    addressable (2.8.7)
+      public_suffix (>= 2.0.2, < 7.0)
     ast (2.4.2)
+    bigdecimal (3.1.8)
+    canister (0.9.2)
+    climate_control (1.2.0)
     coderay (1.1.3)
+    crack (1.0.0)
+      bigdecimal
+      rexml
     diff-lcs (1.5.1)
     docile (1.4.1)
-    json (2.7.2)
+    dotenv (3.1.6)
+    faraday (2.12.2)
+      faraday-net_http (>= 2.0, < 3.5)
+      json
+      logger
+    faraday-net_http (3.4.0)
+      net-http (>= 0.5.0)
+    hashdiff (1.1.2)
+    json (2.9.0)
     language_server-protocol (3.17.0.3)
     lint_roller (1.1.0)
+    logger (1.6.3)
     method_source (1.1.0)
+    mysql2 (0.5.6)
+    net-http (0.6.0)
+      uri
     parallel (1.26.3)
-    parser (3.3.5.0)
+    parser (3.3.6.0)
       ast (~> 2.4.1)
       racc
-    pry (0.14.2)
+    pry (0.15.0)
       coderay (~> 1.1)
       method_source (~> 1.0)
+    public_suffix (6.0.1)
     racc (1.8.1)
     rainbow (3.1.1)
-    regexp_parser (2.9.2)
-    rexml (3.3.9)
+    regexp_parser (2.9.3)
+    rexml (3.4.0)
     rspec (3.13.0)
       rspec-core (~> 3.13.0)
       rspec-expectations (~> 3.13.0)
       rspec-mocks (~> 3.13.0)
-    rspec-core (3.13.1)
+    rspec-core (3.13.2)
       rspec-support (~> 3.13.0)
     rspec-expectations (3.13.3)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.13.0)
-    rspec-mocks (3.13.1)
+    rspec-mocks (3.13.2)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.13.0)
-    rspec-support (3.13.1)
-    rubocop (1.65.1)
+    rspec-support (3.13.2)
+    rubocop (1.69.2)
       json (~> 2.3)
       language_server-protocol (>= 3.17.0)
       parallel (~> 1.10)
       parser (>= 3.3.0.2)
       rainbow (>= 2.2.2, < 4.0)
-      regexp_parser (>= 2.4, < 3.0)
-      rexml (>= 3.2.5, < 4.0)
-      rubocop-ast (>= 1.31.1, < 2.0)
+      regexp_parser (>= 2.9.3, < 3.0)
+      rubocop-ast (>= 1.36.2, < 2.0)
       ruby-progressbar (~> 1.7)
-      unicode-display_width (>= 2.4.0, < 3.0)
-    rubocop-ast (1.32.3)
+      unicode-display_width (>= 2.4.0, < 4.0)
+    rubocop-ast (1.37.0)
       parser (>= 3.3.1.0)
-    rubocop-performance (1.21.1)
+    rubocop-performance (1.23.0)
       rubocop (>= 1.48.1, < 2.0)
       rubocop-ast (>= 1.31.1, < 2.0)
     ruby-progressbar (1.13.0)
+    sequel (5.87.0)
+      bigdecimal
     simplecov (0.22.0)
       docile (~> 1.1)
       simplecov-html (~> 0.11)
       simplecov_json_formatter (~> 0.1)
     simplecov-html (0.13.1)
     simplecov-lcov (0.8.0)
     simplecov_json_formatter (0.1.4)
-    standard (1.40.0)
+    standard (1.43.0)
       language_server-protocol (~> 3.17.0.2)
       lint_roller (~> 1.0)
-      rubocop (~> 1.65.0)
+      rubocop (~> 1.69.1)
       standard-custom (~> 1.0.0)
-      standard-performance (~> 1.4)
+      standard-performance (~> 1.6)
     standard-custom (1.0.2)
       lint_roller (~> 1.0)
       rubocop (~> 1.50)
-    standard-performance (1.4.0)
+    standard-performance (1.6.0)
       lint_roller (~> 1.1)
-      rubocop-performance (~> 1.21.0)
+      rubocop-performance (~> 1.23.0)
     standardrb (1.0.1)
       standard
-    unicode-display_width (2.6.0)
+    unicode-display_width (3.1.2)
+      unicode-emoji (~> 4.0, >= 4.0.4)
+    unicode-emoji (4.0.4)
+    uri (1.0.2)
+    webmock (3.24.0)
+      addressable (>= 2.8.0)
+      crack (>= 0.3.2)
+      hashdiff (>= 0.4.0, < 2.0.0)
 
 PLATFORMS
   aarch64-linux
   ruby
 
 DEPENDENCIES
+  canister
+  climate_control
+  dotenv
+  faraday
+  mysql2
   pry
   rspec
+  sequel
   simplecov
   simplecov-lcov
   standardrb
+  webmock
 
 BUNDLED WITH
-   2.5.19
+   2.5.23

README.md

@@ -29,15 +29,47 @@ docker compose run --rm test bundle exec standardrb
 docker compose run --rm test bundle exec rspec
 ```
 
-run_process_zephir_incremental.sh (daily)
-=========================================
+## Standard Locations
+
+Post-Zephir can read and write files in a number of locations, and it can become bewildering.
+Many of the locations (all of them directories) show up again and again. Under Argo these
+all come from the `ENV` provided to the workflow. Under Docker the locations are not so scattered,
+and all orient themselves to `ENV[ROOTDIR]`. The shell scripts rely on `config/defaults` to fill
+in many of these variables; the Ruby scripts expect that the environment variables set by `config/defaults` are present.
+
+TODO: can we use `dotenv` and `.env` in both the shell scripts and the Ruby code, and get rid of
+`config/defaults`? Or can we translate `config/defaults` into Ruby and invoke it from the driver?
+
+| `ENV`               | Standard Location                   | Docker/Default Location       |
+| --------            | -------                             | -----                         |
+| `CATALOG_ARCHIVE`   | `/htapps/archive/catalog`           | `DATA_ROOT/catalog_archive`   |
+| `CATALOG_PREP`      | `/htsolr/catalog/prep`              | `DATA_ROOT/catalog_prep`      |
+| `DATA_ROOT`         | `/htprep/zephir`                    | `ROOTDIR/data`                |
+| `FEDDOCS_HOME`      | `/htprep/govdocs`                   | `DATA_ROOT/govdocs`           |
+| `INGEST_BIBRECORDS` | `/htapps/babel/feed/var/bibrecords` | `DATA_ROOT/ingest_bibrecords` |
+| `RIGHTS_DIR`        | `/htapps/babel/feed/var/rights`     | `DATA_ROOT/rights`            |
+| `ROOTDIR`           | (not used)                          | `/usr/src/app`                |
+
+Additional derivative paths are set by `config/defaults`, typically from the daily or monthly shell script.
+
+| `ENV`               | Standard/Default/Docker Location    | Note             |
+| --------            | -------                             | ----             |
+| `REPORTS`           | `DATA_ROOT/reports`                 | *unused*         |
+| `RIGHTS_DBM`        | `DATA_ROOT/rights_dbm`              | *this is a file* |
+| `TMPDIR`            | `DATA_ROOT/work`                    |                  |
+| `ZEPHIR_DATA`       | `DATA_ROOT/zephir`                  |                  |
+
+
+
+## `run_process_zephir_incremental.sh` (daily)
+
 * Process daily file of new/updated/deleted metadata provided by Zephir
-* Send deleted bib record IDs (provided by Zephir) to Bill
-* "Clean up" zephir records
+* Send deleted bib record IDs (provided by Zephir) to catalog indexer
+* "Clean up" zephir records (what does this mean?)
 * (re)determine bibliographic rights
-  + Write new/updated bib rights to file for Aaron's process to pick up and update the rights db (Why: possibly because of limited permissions on the rights database)
-* File of processed new/updated records is copied to an HT server for Bill to index in the catalog
-* Retrieves full bib metadata file from zephir and runs run_zephir_full_monthly.sh. (Why?)
+  + Write new/updated bib rights to file for `populate_rights_data.pl` to pick up and update the rights db
+* File of processed new/updated records is copied to a location for the catalog indexer to find it
+* Retrieves full bib metadata file from zephir and runs `run_zephir_full_monthly.sh`. (It does?? I don't think so.)
 
 Why?
 ----
@@ -47,18 +79,33 @@ Data In
 -------
 * `ht_bib_export_incr_YYYY-MM-DD.json.gz` (incremental updates from Zephir, `ftps_zephir_get`)
 * `vufind_removed_cids_YYYY-MM-DD.txt.gz` (CIDs that have gone away, `ftps_zephir_get`)
-* `/tmp/rights_dbm`  (taken from `ht_rights.rights_current` table in the rights database)
-* `us_cities.db` (dependency for `bib_rights.pm`)
-* `us_fed_pub_exception_file` (dependency for `bib_rights.pm`, `/htdata/govdocs/feddocs_oclc_filter/`) 
+* `DATA_ROOT/rights_dbm` (local copy of Rights DB `ht_rights.rights_current`)
+* `ROOTDIR/data/us_cities.db` (dependency for `bib_rights.pm`)
+* `ENV[us_fed_pub_exception_file]` (optional dependency for `bib_rights.pm`)
 
 Data Out
 --------
-* `debug_current.txt` (what and why for this?)
-* `zephir_upd_YYYYMMDD.rights` - picked up hourly by https://github.com/hathitrust/feed_internal/blob/master/feed.hourly/populate_rights_data.pl and loaded into the `rights_current` table. Will be placed directly in /htapps/babel/feed/var/rights and will remove the scp logic from populate_rights_data.pl
-* `zephir_upd_YYYYMMDD_delete.txt.gz`  will be moved to /htsolr/catalog/prep. Used by the catalog to process deletes.
-* `zephir_upd_YYYYMMDD_dollar_dup.txt `(generated by post_zephir_cleanup.pl, gets sent to Zephir, ftps_zephir_send, Zephir uningests these duplicate records)
-* `zephir_upd_YYYYMMDD.json.gz` will be sent to /htsolr/catalog/prep for [catalog indexing](https://github.com/hathitrust/hathitrust_catalog_indexer)
-* `zephir_full_monthly_rpt.txt` Does anyone need this?
+
+Many files are named based on the `BASENAME` variable which is "zephir_upd_YYYYMMDD." Files are typically created in
+`TMPDIR` and moved/renamed from there.
+
+AFAICT, Verifier should only be interested in files outside `TMPDIR`, with the possible exception of
+`TMPDIR/vufind_incremental_YYYY-MM-DD_dollar_dup.txt.gz`.
+
+| File                                                     | Notes                                                                                        |
+| --------                                                 | -----                                                                                        |
+| `CATALOG_ARCHIVE/zephir_upd_YYYYMMDD.json.gz`            | From `postZephir.pm`: gzipped and copied (not moved) by shell script                         |
+| `CATALOG_PREP/zephir_upd_YYYYMMDD.json.gz`               | Same file as above, removed from `TMPDIR` after being copied to the two destinations         |
+| `CATALOG_PREP/zephir_upd_YYYYMMDD_delete.txt.gz`         | Created as `TMPDIR/BASENAME_all_delete.txt.gz` combining two files (see below)               |
+| `RIGHTS_DIR/zephir_upd_YYYYMMDD.rights`                  | From `postZephir.pm`: moved from `TMPDIR`                                                    |
+| `ROOTDIR/data/zephir/debug_current.txt`                  | _Commented out at end of monthly script. Should be removed._                                 |
+| `TMPDIR/vufind_incremental_YYYY-MM-DD_dollar_dup.txt.gz` | Created as `TMPDIR/BASENAME_dollar_dup.txt`, renamed and sent to Zephir                      |
+| `TMPDIR/zephir_upd_YYYYMMDD_delete.txt`                  | From `postZephir.pm`: usually empty list of 974-less CIDs, merged with `vufind_removed_cids` |
+| `TMPDIR/zephir_upd_YYYYMMDD.rights.debug`                | From `postZephir.pm`, _if no one is using this it should be removed_                         |
+| `TMPDIR/zephir_upd_YYYYMMDD_rpt.txt`                     | Log data from `postZephir.pm`                                                                |
+| `TMPDIR/zephir_upd_YYYYMMDD_stderr`                      | `STDERR` from `postZephir.pm`, _if no one is using this it should be removed_                |
+| `TMPDIR/zephir_upd_YYYYMMDD_zephir_delete.txt`           | Intermediate file from `vufind_removed_cids_...` before merge with our deletes, _remove?_    |
+
 
 Perl script dependencies
 ------------------------
@@ -87,20 +134,28 @@ Previously generated the HTRC datasets. All that remains is the zephir_ingested_
 
 Data In
 -------
+* `ht_bib_export_full_YYYY-MM-DD.json.gz` (monthly updates from Zephir, `ftps_zephir_get`)
+  Note: this file is deleted by the `unpigz` command that splits it into smaller files to process in parallel.
+* Note: there is no monthly "removed CIDs" or "deletes" files, these are only in the daily updates.
 * US Fed Doc exception list `/htdata/govdocs/feddocs_oclc_filter/oclcs_removed_from_registry.txt`
-* `/tmp/rights_dbm`
+* `DATA_ROOT/rights_dbm` (local copy of Rights DB `ht_rights.rights_current`)
 * `groove_export_YYYY-MM-DD.tsv.gz` (ftps from cdlib)
-* `ht_bib_export_full_YYYY-MM-DD.json.gz`
+
 
 Data Out
 --------
-* `groove_export_YYYY-MM-DD.tsv.gz` will be moved to /htapps/babel/feed/var/bibrecords/groove_full.tsv.gz  
-* `zephir_full_${YESTERDAY}_vufind.json.gz` catalog archive. Indexed into catalog via the same process as for `run_process_zephir_incremental.sh`
-* `zephir_full_${YESTERDAY}.rights` moved to /htapps/babel/feed/var/rights/
-* `zephir_full_${YESTERDAY}.rights.debug`, doesn't appear to be used
-* `zephir_full_monthly_rpt.txt`moved to ../data/full/
-* `zephir_full_${YESTERDAY}.rights_rpt.tsv moved to ./data/full/
-* `zephir_ingested_items.txt.gz` - copied to `/htapps/babel/feed/var/bibrecords`. Used by https://github.com/hathitrust/feed_internal/blob/master/feed.monthly/zephir_diff.pl to refresh the full `feed_zephir_items` table on a monthly basis.
+| File                                                     | Notes                                                                                        |
+| --------                                                 | -----                                                                                        |
+| `INGEST_BIBRECORDS/groove_full.tsv.gz`                   | Downloaded as `groove_export_YYYY-MM-DD.tsv.gz` and moved, contents are not modified         |
+| `INGEST_BIBRECORDS/zephir_ingested_items.txt.gz`         | From `postZephir.pm`, TSV of {htid, source, collection, digitization_source, ia_id}          |
+| `CATALOG_ARCHIVE/zephir_full_YYYYMMDD_vufind.json.gz`    | Concatenated from parallel-processed files, gzipped and moved by shell script                |
+| `CATALOG_PREP/zephir_full_YYYYMMDD_vufind.json.gz`       | Same file as above, copied to `CATALOG_PREP` before being moved to `CATALOG_ARCHIVE`         |
+| `RIGHTS_DIR/zephir_full_YYYYMMDD.rights`                 | From `postZephir.pm`: moved from `TMPDIR`                                                    |
+| `TMPDIR/stderr.tmp.txt`                                  | Concatenated from subfiles' STDERR                                                           |
+| `TMPDIR/zephir_full_YYYYMMDD.rights.debug`               | From `postZephir.pm`, _if no one is using this it should be removed_                         |                                                                       |
+| `ZEPHIR_DATA/full/zephir_full_monthly_rpt.txt`           | Concatenated from subfiles and moved from `TMPDIR`                                           |
+| `ZEPHIR_DATA/full/zephir_full_YYYYMMDD.rights_rpt.tsv`   | Concatenated from subfiles and moved from `TMPDIR`                                           |
+
 
 Perl script dependencies
 ------------------------

bin/post_zephir.rb

@@ -9,6 +9,7 @@
 
 require_relative "../lib/dates"
 require_relative "../lib/derivatives"
+require_relative "../lib/journal"
 
 def run_system_command(command)
   LOGGER.info command
@@ -22,18 +23,24 @@ def run_system_command(command)
 YESTERDAY = Date.today - 1
 
 inventory = PostZephirProcessing::Derivatives.new(date: YESTERDAY)
-
-if inventory.earliest_missing_date.nil?
-  LOGGER.info "no Zephir files to process, exiting"
-  exit 0
+dates = []
+# Is there a missing date? Plug them into an array to process.
+if !inventory.earliest_missing_date.nil?
+  dates = (inventory.earliest_missing_date..YESTERDAY)
 end
 
-dates = (inventory.earliest_missing_date..YESTERDAY)
 LOGGER.info "Processing Zephir files from #{dates}"
 dates.each do |date|
   date_str = date.strftime("%Y%m%d")
+  LOGGER.info "Processing Zephir file for #{date_str}"
   if date.last_of_month?
     run_system_command "#{FULL_SCRIPT} #{date_str}"
   end
   run_system_command "#{INCREMENTAL_SCRIPT} #{date_str}"
 end
+
+# Record our work for the verifier
+LOGGER.info "Writing journal for #{dates}"
+# TODO: consider moving the `to_a` to the Journal initializer so it can take
+# Ranges as well as Arrays
+PostZephirProcessing::Journal.new(dates: dates.to_a).write!