Functionality to accept compressed files as input to predict when using a Predictor

allennlp/commands/predict.py

@@ -13,7 +13,7 @@
 from allennlp.commands.subcommand import Subcommand
 from allennlp.common import logging as common_logging
 from allennlp.common.checks import check_for_gpu, ConfigurationError
-from allennlp.common.file_utils import cached_path
+from allennlp.common.file_utils import cached_path, open_compressed
 from allennlp.common.util import lazy_groups_of
 from allennlp.data.dataset_readers import MultiTaskDatasetReader
 from allennlp.models.archival import load_archive
@@ -73,6 +73,14 @@ def add_subparser(self, parser: argparse._SubParsersAction) -> argparse.Argument
             "flag is set.",
         )
 
+        subparser.add_argument(
+            "--compression-type",
+            type=str,
+            choices=["gz", "bz2", "lzma"],
+            default=None,
+            help="Indicates the compressed format of the input file.",
+        )
+
         subparser.add_argument(
             "--multitask-head",
             type=str,
@@ -152,6 +160,7 @@ def __init__(
         batch_size: int,
         print_to_console: bool,
         has_dataset_reader: bool,
+        compression_type: str = None,
         multitask_head: Optional[str] = None,
     ) -> None:
         self._predictor = predictor
@@ -160,7 +169,7 @@ def __init__(
         self._batch_size = batch_size
         self._print_to_console = print_to_console
         self._dataset_reader = None if not has_dataset_reader else predictor._dataset_reader
-
+        self.compression_type = compression_type
         self._multitask_head = multitask_head
         if self._multitask_head is not None:
             if self._dataset_reader is None:
@@ -212,10 +221,21 @@ def _get_json_data(self) -> Iterator[JsonDict]:
                     yield self._predictor.load_line(line)
         else:
             input_file = cached_path(self._input_file)
-            with open(input_file, "r") as file_input:
-                for line in file_input:
-                    if not line.isspace():
-                        yield self._predictor.load_line(line)
+            try:
+                with open_compressed(input_file) as file_input:
+                    for line in file_input:
+                        if not line.isspace():
+                            yield self._predictor.load_line(line)
+            except OSError:
+                if self.compression_type:
+                    with open_compressed(input_file, self.compression_type) as file_input:
+                        for line in file_input:
+                            if not line.isspace():
+                                yield self._predictor.load_line(line)
+                else:
+                    print(
+                        "Automatic detection of compression type failed, please specify the compression type argument"
+                    )
 
     def _get_instance_data(self) -> Iterator[Instance]:
         if self._input_file == "-":

allennlp/common/file_utils.py

@@ -1085,20 +1085,27 @@ def get_file_extension(path: str, dot=True, lower: bool = True):
 
 
 def open_compressed(
-    filename: Union[str, PathLike], mode: str = "rt", encoding: Optional[str] = "UTF-8", **kwargs
+    filename: Union[str, PathLike],
+    compression_type: str = None,
+    mode: str = "rt",
+    encoding: Optional[str] = "UTF-8",
+    **kwargs,
 ):
     if not isinstance(filename, str):
         filename = str(filename)
     open_fn: Callable = open
 
-    if filename.endswith(".gz"):
-        import gzip
-
-        open_fn = gzip.open
-    elif filename.endswith(".bz2"):
-        import bz2
+    compression_modules = {"gz": "gzip", "bz2": "bz2", "lzma": "lzma"}
+    if not compression_type:
+        for extension in compression_modules:
+            if filename.endswith(extension):
+                module = __import__(compression_modules[extension])
+                open_fn = module.open
+                break 
+    else:
+        module = __import__(compression_modules[extension])
+        open_fn = module.open
 
-        open_fn = bz2.open
     return open_fn(cached_path(filename), mode=mode, encoding=encoding, **kwargs)
 
 

tests/common/file_utils_test.py

@@ -347,7 +347,7 @@ def test_open_compressed(self):
         with open_compressed(uncompressed_file) as f:
             uncompressed_lines = [line.strip() for line in f]
 
-        for suffix in ["bz2", "gz"]:
+        for suffix in ["bz2", "gz", "lzma"]:
             compressed_file = f"{uncompressed_file}.{suffix}"
             with open_compressed(compressed_file) as f:
                 compressed_lines = [line.strip() for line in f]