Exit with cleaner message when no entries are expected in the ZIM

CHANGELOG.md

@@ -11,6 +11,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 - Moved rules definition from JSON to YAML and documented update process (#216)
 
+### Added
+
+- Exit with cleaner message when no entries are expected in the ZIM (#336) and when main entry is not processable (#337)
+
 ## [2.0.2] - 2024-06-18
 
 ### Added

src/warc2zim/converter.py

@@ -92,6 +92,12 @@
 PY2JS_RULE_RX = re.compile(r"\\(\d)", re.ASCII)
 
 
+class UnprocessableWarcError(Exception):
+    """Exception raised when it is not possible to process WARC file(s) received"""
+
+    ...
+
+
 class Converter:
     def __init__(self, args):
         if args.verbose:
@@ -262,7 +268,12 @@ def run(self):
             )
             return 100
 
-        self.gather_information_from_warc()
+        try:
+            self.gather_information_from_warc()
+        except UnprocessableWarcError as exc:
+            logger.error(exc)
+            return 4
+
         # Fallback language
         if not self.language:
             logger.warning("No valid ZIM language, fallbacking to `eng`.")
@@ -391,6 +402,10 @@ def gather_information_from_warc(self):
 
             status_code = get_status_code(record)
             if not can_process_status_code(status_code):
+                if record.rec_type == "response" and self.main_path == zim_path:
+                    raise UnprocessableWarcError(
+                        f"Main URL returned an unprocessable HTTP code: {status_code}"
+                    )
                 continue
 
             if status_code_is_processable_redirect(status_code):
@@ -480,10 +495,16 @@ def gather_information_from_warc(self):
             logger.debug(f"Favicon: {self.favicon_url or self.favicon_path}")
             main_page_found = True
 
+        if len(self.expected_zim_items) == 0:
+            raise UnprocessableWarcError(
+                "No entry found to push to the ZIM, WARC file(s) is unprocessable "
+                "and looks probably mostly empty"
+            )
+
         logger.info(f"Expecting {len(self.expected_zim_items)} ZIM entries to files")
 
         if not main_page_found:
-            raise KeyError(
+            raise UnprocessableWarcError(
                 f"Unable to find WARC record for main page: {self.main_path}, aborting"
             )
 

tests/test_warc_to_zim.py

@@ -4,6 +4,7 @@
 import io
 import json
 import os
+import pathlib
 import re
 import time
 from urllib.parse import unquote
@@ -607,7 +608,7 @@ def test_fuzzy_urls(self, tmp_path, fuzzycheck):
 
     def test_error_bad_main_page(self, tmp_path):
         zim_output_not_created = "zim-out-not-created.zim"
-        with pytest.raises(KeyError, match="Unable to find WARC record for main page:"):
+        assert (
             main(
                 [
                     "-v",
@@ -622,6 +623,29 @@ def test_error_bad_main_page(self, tmp_path):
                     zim_output_not_created,
                 ]
             )
+            == 4
+        )
+
+    def test_error_main_page_unprocessable(self, tmp_path):
+        zim_output_not_created = "zim-out-not-created.zim"
+        assert (
+            main(
+                [
+                    "-v",
+                    os.path.join(TEST_DATA_DIR, "main-entry-403.warc.gz"),
+                    "-u",
+                    "https://wikizilla.org/wiki/Doug",
+                    "--output",
+                    str(tmp_path),
+                    "--name",
+                    "bad",
+                    "--zim-file",
+                    zim_output_not_created,
+                ]
+            )
+            == 4
+        )
+        assert not (pathlib.Path(tmp_path) / zim_output_not_created).exists()
 
     def test_args_only(self):
         # error, name required