Friendlier error when lzma is not there and importing trafilatura fails

Fix #937

minet/cli/argparse.py

@@ -293,7 +293,14 @@ def build_parser(name, version, commands):
 
     return parser, subparser_index
 
-FOLDER_STRATEGY_CHOICES = ["flat", "fullpath", "hostname", "normalize-hostname", "prefix-x"]
+
+FOLDER_STRATEGY_CHOICES = [
+    "flat",
+    "fullpath",
+    "hostname",
+    "normalize-hostname",
+    "prefix-x",
+]
 
 # NOTE: indentation IS important
 FOLDER_STRATEGY_DOCUMENTATION = """
@@ -321,6 +328,7 @@ def build_parser(name, version, commands):
             "fr.", for instance) and their public suffix will be dropped.
 """
 
+
 class FolderStrategyType:
     def __call__(self, name):
         from minet.fs import FolderStrategy
@@ -601,7 +609,9 @@ def __init__(
         **kwargs,
     ):
         fields_help = (
-            "Available flags are: " + and_join([f"`{f}`" for f in TRAFILATURA_FIELDNAMES]) + "."
+            "Available flags are: "
+            + and_join([f"`{f}`" for f in TRAFILATURA_FIELDNAMES])
+            + "."
         )
         help = fields_help if not help else help + " " + fields_help
         super().__init__(option_strings, dest, help=help, default=default, **kwargs)

minet/cli/crawl/__init__.py

@@ -6,7 +6,8 @@
     command,
     FolderStrategyType,
     BooleanAction,
-    ExtractionSelectionAction,FOLDER_STRATEGY_DOCUMENTATION
+    ExtractionSelectionAction,
+    FOLDER_STRATEGY_DOCUMENTATION,
 )
 from minet.cli.exceptions import InvalidArgumentsError
 

minet/cli/fetch/__init__.py

@@ -1,6 +1,10 @@
 from casanova import IndexedResumer
 
-from minet.cli.argparse import command, FolderStrategyType, FOLDER_STRATEGY_DOCUMENTATION
+from minet.cli.argparse import (
+    command,
+    FolderStrategyType,
+    FOLDER_STRATEGY_DOCUMENTATION,
+)
 from minet.cli.constants import DEFAULT_CONTENT_FOLDER, DEFAULT_SCREENSHOT_FOLDER
 from minet.cli.exceptions import InvalidArgumentsError
 

minet/extraction.py

@@ -2,7 +2,16 @@
 
 from dataclasses import dataclass, field
 from casanova import TabularRecord
-from trafilatura.core import bare_extraction
+
+try:
+    from trafilatura.core import bare_extraction
+except ModuleNotFoundError as e:
+    if "lzma" in str(e):
+        raise ImportError(
+            "cannot import trafilatura because your version of python was not compiled with lzma.\nSee https://stackoverflow.com/questions/57743230/userwarning-could-not-import-the-lzma-module-your-installed-python-is-incomple for potential solutions."
+        )
+
+    raise
 
 from minet.exceptions import TrafilaturaError
 from minet.encodings import fix_surrogates