Apply Sourcery suggestions and fix typos

Author: DimitriPapadopoulos

docs/source/changelog.rst

@@ -103,7 +103,7 @@ Enhancements
 - "tree" text display of filesystem contents (#1750)
 - async wrapper for sync FSs (#1745)
 - new known implementation: tosfs (#1739)
-- consilidate block fetch requests (#1733)
+- consolidate block fetch requests (#1733)
 
 Fixes
 

fsspec/archive.py

@@ -43,8 +43,8 @@ def info(self, path, **kwargs):
             return {"name": "", "type": "directory", "size": 0}
         if path in self.dir_cache:
             return self.dir_cache[path]
-        elif path + "/" in self.dir_cache:
-            return self.dir_cache[path + "/"]
+        elif f"{path}/" in self.dir_cache:
+            return self.dir_cache[f"{path}/"]
         else:
             raise FileNotFoundError(path)
 
@@ -69,7 +69,6 @@ def ls(self, path, detail=True, **kwargs):
                     out = {"name": ppath, "size": 0, "type": "directory"}
                     paths[ppath] = out
         if detail:
-            out = sorted(paths.values(), key=operator.itemgetter("name"))
-            return out
+            return sorted(paths.values(), key=operator.itemgetter("name"))
         else:
             return sorted(paths)

fsspec/asyn.py

@@ -85,10 +85,8 @@ def sync(loop, func, *args, timeout=None, **kwargs):
     result = [None]
     event = threading.Event()
     asyncio.run_coroutine_threadsafe(_runner(event, coro, result, timeout), loop)
-    while True:
-        # this loops allows thread to get interrupted
-        if event.wait(1):
-            break
+    while not event.wait(1):
+        # this loop allows thread to get interrupted
         if timeout is not None:
             timeout -= 1
             if timeout < 0:
@@ -356,10 +354,11 @@ async def _copy(
         batch_size=None,
         **kwargs,
     ):
-        if on_error is None and recursive:
-            on_error = "ignore"
-        elif on_error is None:
-            on_error = "raise"
+        if on_error is None:
+            if recursive:
+                on_error = "ignore"
+            else:
+                on_error = "raise"
 
         if isinstance(path1, list) and isinstance(path2, list):
             # No need to expand paths when both source and destination
@@ -714,7 +713,7 @@ async def _walk(self, path, maxdepth=None, on_error="omit", **kwargs):
         detail = kwargs.pop("detail", False)
         try:
             listing = await self._ls(path, detail=True, **kwargs)
-        except (FileNotFoundError, OSError) as e:
+        except OSError as e:
             if on_error == "raise":
                 raise
             elif callable(on_error):
@@ -766,7 +765,7 @@ async def _glob(self, path, maxdepth=None, **kwargs):
         ends_with_sep = path.endswith(seps)  # _strip_protocol strips trailing slash
         path = self._strip_protocol(path)
         append_slash_to_dirname = ends_with_sep or path.endswith(
-            tuple(sep + "**" for sep in seps)
+            tuple(f"{sep}**" for sep in seps)
         )
         idx_star = path.find("*") if path.find("*") >= 0 else len(path)
         idx_qmark = path.find("?") if path.find("?") >= 0 else len(path)
@@ -814,7 +813,7 @@ async def _glob(self, path, maxdepth=None, **kwargs):
             p: info
             for p, info in sorted(allpaths.items())
             if pattern.match(
-                p + "/"
+                f"{p}/"
                 if append_slash_to_dirname and info["type"] == "directory"
                 else p
             )

fsspec/caching.py

@@ -430,8 +430,7 @@ def _fetch_block(self, block_number: int) -> bytes:
         self.total_requested_bytes += end - start
         self.miss_count += 1
         logger.info("BlockCache fetching block %d", block_number)
-        block_contents = super()._fetch(start, end)
-        return block_contents
+        return super()._fetch(start, end)
 
     def _read_cache(
         self, start: int, end: int, start_block_number: int, end_block_number: int
@@ -704,7 +703,7 @@ class UpdatableLRU(Generic[P, T]):
     """
     Custom implementation of LRU cache that allows updating keys
 
-    Used by BackgroudBlockCache
+    Used by BackgroundBlockCache
     """
 
     class CacheInfo(NamedTuple):
@@ -857,7 +856,7 @@ def _fetch(self, start: int | None, end: int | None) -> bytes:
                     self._fetch_future = None
                 else:
                     # Must join if we need the block for the current fetch
-                    must_join = bool(
+                    must_join = (
                         start_block_number
                         <= self._fetch_future_block_number
                         <= end_block_number
@@ -920,8 +919,7 @@ def _fetch_block(self, block_number: int, log_info: str = "sync") -> bytes:
         logger.info("BlockCache fetching block (%s) %d", log_info, block_number)
         self.total_requested_bytes += end - start
         self.miss_count += 1
-        block_contents = super()._fetch(start, end)
-        return block_contents
+        return super()._fetch(start, end)
 
     def _read_cache(
         self, start: int, end: int, start_block_number: int, end_block_number: int

fsspec/core.py

@@ -339,7 +339,7 @@ def _un_chain(path, kwargs):
             if "://" in p or x.match(p):
                 bits.append(p)
             else:
-                bits.append(p + "://")
+                bits.append(f"{p}://")
     else:
         bits = [path]
     # [[url, protocol, kwargs], ...]

fsspec/fuse.py

@@ -76,8 +76,7 @@ def read(self, path, size, offset, fh):
 
         f = self.cache[fh]
         f.seek(offset)
-        out = f.read(size)
-        return out
+        return f.read(size)
 
     def write(self, path, data, offset, fh):
         logger.debug("write %s", (path, offset))
@@ -119,7 +118,7 @@ def unlink(self, path):
         fn = "".join([self.root, path.lstrip("/")])
         try:
             self.fs.rm(fn, False)
-        except (OSError, FileNotFoundError) as exc:
+        except OSError as exc:
             raise FuseOSError(EIO) from exc
 
     def release(self, path, fh):

fsspec/implementations/cache_metadata.py

@@ -166,7 +166,7 @@ def on_close_cached_file(self, f: Any, path: str) -> None:
 
         The actual closing of the file is the responsibility of the caller.
         """
-        # File must be writeble, so in self.cached_files[-1]
+        # File must be writeable, so in self.cached_files[-1]
         c = self.cached_files[-1][path]
         if c["blocks"] is not True and len(c["blocks"]) * f.blocksize >= f.size:
             c["blocks"] = True

fsspec/implementations/cached.py

@@ -338,7 +338,7 @@ def _open(
         # explicitly submitting the size to the open call will avoid extra
         # operations when opening. This is particularly relevant
         # for any file that is read over a network, e.g. S3.
-        size = detail.get("size", None)
+        size = detail.get("size")
 
         # call target filesystems open
         self._mkcache()
@@ -821,7 +821,7 @@ def info(self, path, **kwargs):
             if f:
                 size = os.path.getsize(f[0].fn) if f[0].closed else f[0].tell()
                 return {"name": path, "size": size, "type": "file"}
-            f = any(_.path.startswith(path + "/") for _ in self.transaction.files)
+            f = any(_.path.startswith(f"{path}/") for _ in self.transaction.files)
             if f:
                 return {"name": path, "size": 0, "type": "directory"}
         return self.fs.info(path, **kwargs)

fsspec/implementations/github.py

@@ -153,7 +153,7 @@ def ls(self, path, detail=False, sha=None, _sha=None, **kwargs):
             _sha = sha or self.root
             for part in parts:
                 out = self.ls(so_far, True, sha=sha, _sha=_sha)
-                so_far += "/" + part if so_far else part
+                so_far += f"/{part}" if so_far else part
                 out = [o for o in out if o["name"] == so_far]
                 if not out:
                     raise FileNotFoundError(path)

fsspec/implementations/http.py

@@ -446,7 +446,7 @@ async def _glob(self, path, maxdepth=None, **kwargs):
         """
         Find files by glob-matching.
 
-        This implementation is idntical to the one in AbstractFileSystem,
+        This implementation is identical to the one in AbstractFileSystem,
         but "?" is not considered as a character for globbing, because it is
         so common in URLs, often identifying the "query" part.
         """
@@ -614,9 +614,11 @@ def read(self, length=-1):
             read only part of the data will raise a ValueError.
         """
         if (
-            (length < 0 and self.loc == 0)  # explicit read all
+            # explicit read all
+            length < 0
+            and self.loc == 0
             # but not when the size is known and fits into a block anyways
-            and not (self.size is not None and self.size <= self.blocksize)
+            and (self.size is None or self.size > self.blocksize)
         ):
             self._fetch_all()
         if self.size is None: