squirrel cache

squirrel cache into node_modules/ which is cached by default across builds

docs/data/presse.parquet.sh

@@ -1,30 +1,29 @@
 # Use "eleventy" .cache to store our temp files
-export TMPDIR=".cache"
+export TMPDIR="docs/.observablehq/.cache"
 mkdir -p $TMPDIR
 
-if [ ! -f "$TMPDIR/presse.parquet" ]; then
+echo "running loader" >&2
 
-  # install duckdb if not already present
-  export PATH=.cache:$PATH
-  command -v duckdb || $(
-    curl --location --output duckdb.zip \
-      https://github.com/duckdb/duckdb/releases/download/v0.10.0/duckdb_cli-linux-amd64.zip && \
-      unzip -qq duckdb.zip && chmod +x duckdb && mkdir -p .cache && mv duckdb .cache/
-  )
+# install duckdb if not already present
+export PATH=$TMPDIR:$PATH
+command -v duckdb || $(
+  curl --location --output duckdb.zip \
+    https://github.com/duckdb/duckdb/releases/download/v0.10.0/duckdb_cli-linux-amd64.zip && \
+    unzip -qq duckdb.zip && chmod +x duckdb && mv duckdb $TMPDIR/
+)
 
-  echo """
-  CREATE TABLE presse AS (
-  SELECT title
-       , author
-       , LPAD((REGEXP_EXTRACT(date, '1[0-9][0-9][0-9]') || '-01-01'), 10, '0')::DATE AS year
-    FROM read_parquet(
-      [('https://huggingface.co/datasets/PleIAs/French-PD-Newspapers/resolve/main/gallica_presse_{:d}.parquet').format(n) for n in range(1, 321)])
-    ORDER BY title, author, year
-  );
+echo """
+CREATE TABLE presse AS (
+SELECT title
+     , author
+     , LPAD((REGEXP_EXTRACT(date, '1[0-9][0-9][0-9]') || '-01-01'), 10, '0')::DATE AS year
+  FROM read_parquet(
+    [('https://huggingface.co/datasets/PleIAs/French-PD-Newspapers/resolve/main/gallica_presse_{:d}.parquet').format(n) for n in range(1, 321)])
+  ORDER BY title, author, year
+);
 
-  COPY presse TO '$TMPDIR/presse.parquet' (COMPRESSION 'ZSTD', row_group_size 10000000);
-  """ | duckdb
-fi
+COPY presse TO '$TMPDIR/presse.parquet' (COMPRESSION 'ZSTD', row_group_size 10000000);
+""" | duckdb
 
 # isatty
 if [ -t 1 ]; then
@@ -32,5 +31,5 @@ if [ -t 1 ]; then
   echo "duckdb -csv :memory: \"SELECT * FROM '$TMPDIR/presse.parquet'\""
 else
   cat $TMPDIR/presse.parquet
-  #rm $TMPDIR/presse.parquet
+  rm $TMPDIR/presse.parquet
 fi 

vercel.json

@@ -1,3 +1,6 @@
 {
-  "cleanUrls": true
+  "buildCommand": "mkdir -p node_modules/cache && ln -s ../node_modules/cache docs/.observablehq && rm -rf dist && observable build",
+  "outputDirectory": "dist",
+  "cleanUrls": true,
+  "framework": null
 }