Master merge conflict resolve

.github/workflows/tests.yml

@@ -0,0 +1,106 @@
+# run test suites
+
+name: Tests
+on:
+  - pull_request
+  - push
+
+jobs:
+  # see: https://github.com/fkirc/skip-duplicate-actions
+  skip_duplicate:
+    continue-on-error: true
+    runs-on: ubuntu-latest
+    outputs:
+      should_skip: ${{ steps.skip_check.outputs.should_skip }}
+    steps:
+      - id: skip_check
+        uses: fkirc/skip-duplicate-actions@master
+        with:
+          concurrent_skipping: "same_content"
+          skip_after_successful_duplicate: "true"
+          do_not_skip: '["pull_request", "workflow_dispatch", "schedule"]'
+
+  # NOTE:
+  #   Run all the steps even if there are no tests defined for a given domain sub-directory.
+  #   This is to make sure that the environment definition is at the very least buildable.
+  tests:
+    needs: skip_duplicate
+    if: ${{ needs.skip_duplicate.outputs.should_skip != 'true' }}
+    runs-on: ${{ matrix.os }}
+    continue-on-error: ${{ matrix.allow-failure }}
+    env:
+      CACHE_NUMBER: 0  # increment to reset cache
+
+    # ensure conda env activation is performed automatically
+    defaults:
+      run:
+        shell: bash -el {0}
+
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest]
+        # somehow mamba with python 3.12 doesn't resolve spacy although available...
+        python-version: ["3.9", "3.10", "3.11"]
+        allow-failure: [false]
+        domain: ["eo", "nlp"]
+
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          fetch-depth: "0"
+
+      - name: Setup Mamba
+        uses: conda-incubator/setup-miniconda@v3
+        with:
+            auto-update-conda: true
+            python-version: ${{ matrix.python-version }}
+            miniforge-variant: Mambaforge
+            miniforge-version: latest
+            activate-environment: github-ci-test-python${{ matrix.python-version }}-${{ matrix.domain }}
+            use-mamba: true
+            use-only-tar-bz2: true
+
+      - name: Set cache date
+        run: echo "DATE=$(date +'%Y%m%d')" >> $GITHUB_ENV
+
+      - uses: actions/cache@v2
+        id: cache
+        with:
+          path: ${{ env.CONDA }}/envs/github-ci-test-python${{ matrix.python-version }}-${{ matrix.domain }}
+          key: conda-python${{ matrix.python-version }}-${{ matrix.domain }}-${{ hashFiles('${{ matrix.domain }}/environment.yml') }}-${{ env.DATE }}-${{ env.CACHE_NUMBER }}
+
+      - name: Display Python
+        run: which python
+
+      - name: Update environment
+        if: steps.cache.outputs.cache-hit != 'true'
+        run: |
+          echo "python=${{ matrix.python-version }}" > ${{ env.CONDA }}/envs/github-ci-test-python${{ matrix.python-version }}-${{ matrix.domain }}/conda-meta/pinned
+          mamba env update \
+            -n github-ci-test-python${{ matrix.python-version }}-${{ matrix.domain }} \
+            -f ${{ matrix.domain }}/environment.yml
+
+      - name: Display Packages
+        if: ${{ matrix.python-version != 'none' }}
+        run: pip freeze
+
+      - name: Display Environment Variables
+        run: |
+          hash -r
+          env | sort
+
+      - name: Check Tests
+        id: check_tests
+        run: |
+          echo "HAS_TEST_DIR=$(test -d ${{ matrix.domain }}/tests && echo 'true' || echo 'false')" >> $GITHUB_OUTPUT
+
+      - name: Install Tests Dependencies
+        if: ${{ steps.check_tests.outputs.HAS_TEST_DIR == 'true' }}
+        run: pip install -r requirements-dev.txt
+
+      - name: Run Tests
+        if: ${{ steps.check_tests.outputs.HAS_TEST_DIR == 'true' }}
+        run: |
+          cd ${{ matrix.domain }}/notebooks
+          python -m pytest -vvv ../tests

.gitignore

@@ -1,8 +1,28 @@
 ### IDE
 **/.idea/
 **/.vscode/
+**/*.code-workspace
+
+### Caches
+**/__pycache__/
+**/cache/
+**/*tree-tagger-linux*
+**/.pytest_cache
+**/condaenv.*.requirements.txt
+
+## Chroma VDB caches
+**/*.bin
+**/*.pickle
+**/*.sqlite3
+
+### Binaries
+**/*.jar
 
 ### Notebooks
 # expect examples per domain
+# disallow notebooks at root
 ./*.ipynb
 **/.ipynb_checkpoints/
+
+### Outputs
+*.log

Makefile

@@ -0,0 +1,18 @@
+# Included custom configs change the value of MAKEFILE_LIST
+# Extract the required reference beforehand so we can use it for help target
+MAKEFILE_NAME := $(word $(words $(MAKEFILE_LIST)),$(MAKEFILE_LIST))
+# Include custom config if it is available
+-include Makefile.config
+
+# Application
+APP_ROOT    := $(abspath $(lastword $(MAKEFILE_NAME))/..)
+APP_NAME    := $(shell basename $(APP_ROOT))
+APP_DOMAINS ?= eo nlp
+DOCKER_REPO ?= crim-ca/pavics-jupyter-images
+
+DOCKER_BUILDS := $(addprefix docker-build-, $(APP_DOMAINS))
+$(DOCKER_BUILDS): docker-build-%:
+	docker build -t $(DOCKER_REPO)/$*:latest "$(APP_ROOT)/$(*)" 2>&1 | tee "$(APP_ROOT)/make-$@.log"
+
+.PHONY: docker-build
+docker-build: $(DOCKER_BUILDS)

eo/environment.yml

@@ -9,7 +9,6 @@ dependencies:
   - intake-stac
   - pyproj
   - rasterio
-  - sat-search
   - shapely
 
   # TODO: These next packages could possibly be added to a more generic 'vision' image, from which 'eo' would be built

nlp/CHANGELOG.md

@@ -3,16 +3,22 @@ Changes
 
 Unreleased (latest)
 ===================
-0.4.1 (2023-11-24)
+
+- n/a
+
+0.5.0 (2023-12-13)
 ===================
+
 Changes:
 --------
+- Update NLU demo notebook with latest pipeline improvements and a STAC wrapper to convert NL queries to STAC requests.
+- Add `duckling` installation in the Docker to allow running it as child process rather than sibling Docker service.
 - Update base image version in Dockerfile
 - Add `mamba` cache cleanup to reduce image size
 
 Fixes:
 ------
-- ...
+- Fix dependencies to make them functional across multiple Python versions.
 
 0.4.0 (2022-12-23)
 ===================
@@ -127,4 +133,4 @@ Changes:
 
 Fixes:
 ------
-- na
+- na

nlp/Dockerfile

@@ -2,14 +2,16 @@ FROM birdhouse/pavics-jupyter-base:0.5.0
 
 # must update conda env as root, because of a permission error when having pip dependencies in the input yml file
 USER root
+# install pre-requirement for compiling some dependencies
+RUN apt install -y pkg-config libpcre++-dev
 
 COPY environment.yml /environment.yml
-COPY notebook_config.yml /notebook_config.yml
 
 # update env "birdy"
 # use umask 0000 so that package files for the updated environment are usable by the user for the jupyter-conda-extension
-RUN umask 0000 && mamba env update -f /environment.yml \
-    && mamba clean -a
+RUN umask 0000 && \
+    mamba env update -f /environment.yml && \
+    mamba clean -y -a
 
 # Set the encoding to UTF-8, this is needed for heideltime to work properly
 ENV LANG=C.UTF-8
@@ -19,21 +21,27 @@ RUN python -m spacy download en_core_web_trf
 
 # Downloading the ner-large flair model
 RUN mkdir flair_models && \
-    curl -L -o flair_models/ner-large  https://huggingface.co/flair/ner-english-large/resolve/main/pytorch_model.bin > /dev/null
+    curl -L -o flair_models/ner-large \
+        https://huggingface.co/flair/ner-english-large/resolve/main/pytorch_model.bin > /dev/null
 
 # Heideltime Tree-tagger Installation
 RUN mkdir -p heideltime/tree-tagger-linux-3.2.3 && cd heideltime/tree-tagger-linux-3.2.3 && \
-    curl -o tree-tagger-linux-3.2.3.tar.gz  https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tree-tagger-linux-3.2.3.tar.gz && \
-    curl -o tagger-scripts.tar.gz https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tagger-scripts.tar.gz && \
-    curl -o english.par.gz https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/english.par.gz && \
-    curl -o install-tagger.sh https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/install-tagger.sh && \
+    curl -o tree-tagger-linux-3.2.3.tar.gz \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tree-tagger-linux-3.2.3.tar.gz && \
+    curl -o tagger-scripts.tar.gz \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tagger-scripts.tar.gz && \
+    curl -o english.par.gz \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/english.par.gz && \
+    curl -o install-tagger.sh \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/install-tagger.sh && \
     tar -xvzf tree-tagger-linux-3.2.3.tar.gz  && \
     sh install-tagger.sh && \
     rm tree-tagger-linux-3.2.3.tar.gz tagger-scripts.tar.gz english.par.gz install-tagger.sh
 
 # Download and extract heideltime.standalone.jar
-RUN cd heideltime && \   
-    curl -L -o heideltime-standalone-2.2.1.tar.gz https://github.com/HeidelTime/heideltime/releases/download/VERSION2.2.1/heideltime-standalone-2.2.1.tar.gz && \
+RUN cd heideltime && \
+    curl -L -o heideltime-standalone-2.2.1.tar.gz \
+        https://github.com/HeidelTime/heideltime/releases/download/VERSION2.2.1/heideltime-standalone-2.2.1.tar.gz && \
     tar -xzvf heideltime-standalone-2.2.1.tar.gz heideltime-standalone/de.unihd.dbs.heideltime.standalone.jar && \
     mv heideltime-standalone/de.unihd.dbs.heideltime.standalone.jar . && \ 
     rmdir heideltime-standalone && \
@@ -42,9 +50,20 @@ RUN cd heideltime && \
 # Give read&write permission to jenkins for config
 RUN chown -R jenkins heideltime
 
+# Setup Haskell for Duckling server
+RUN curl -sSL https://get.haskellstack.org/ | bash && \
+    git clone https://github.com/facebook/duckling && \
+    cd duckling && \
+    stack build && \
+    stack install && \
+    cd .. && \
+    rm -fr duckling
+ENV PATH="/root/.local/bin:$PATH"
+
 # Give ownership of the conda cache folder to jenkins, to enable installing packages by the user from JupyterLab
 RUN mkdir /opt/conda/pkgs/cache && chown -R 1000:1000 /opt/conda/pkgs/cache
 
+COPY notebook_config.yml /notebook_config.yml
+
 # specify user because of problem running start-notebook.sh when being root
 USER jenkins
-

nlp/environment.yml

@@ -5,19 +5,26 @@ channels:
   - conda-forge
 
 dependencies:
-  - intake-esm==2021.1.15
-  - intake-stac==0.3.0
-  - sat-search==0.3.0
+  - intake-esm
+  - intake-stac>=0.4.0
   - threddsclient==0.4.2
-  - openjdk==8.0.152 
-  # python-flair=0.8 only works with numpy<=1.19.5
-  - python-flair=0.8
-  - numpy<=1.19.5
-  - spacy==3.1.0
-  - python-dateutil==2.7.5
-  - python-levenshtein==0.12.2
-  - requests=2.25.1
-  - pip==20.3.3
+  - openjdk==8.0.152
+  - python-flair
+  - numpy
+  - pydantic<2
+  - python-levenshtein
+  - requests
+  - pip>=22
   - pip:
     - textsearch==0.0.21
-    - spacy==3.1.0
+    - osmnx
+    - langchain
+    - spacy>=3.5,<4
+    - spacy-transformers
+    - transformers<4.31
+    - sentence_transformers
+    - chromadb
+    - shapely
+    - ipywidgets
+    - nltk
+    - pystac_client