dvc-utils 0.0.5__tar.gz → 0.0.7__tar.gz

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dvc-utils
-Version: 0.0.5
+Version: 0.0.7
 Summary: CLI for diffing DVC files at two commits (or one commit vs. current worktree), optionally passing both through another command first
 Home-page: https://github.com/runsascoded/dvc-utils
 Author: Ryan Williams
@@ -15,7 +15,7 @@ CLI for diffing [DVC] files, optionally passing both through another command fir
 <!-- toc -->
 - [Installation](#installation)
 - [Usage](#usage)
-    - [`dvc-utils diff`](#dvc-utils-diff)
+    - [`dvc-diff`](#dvc-diff)
 - [Examples](#examples)
     - [Parquet file](#parquet-diff)
         - [Schema diff](#parquet-schema-diff)
@@ -42,11 +42,13 @@ dvc-utils --help
 #         worktree), optionally passing both through another command first
 ```
 
-### `dvc-utils diff` <a id="dvc-utils-diff"></a>
-<!-- `bmdf -- dvc-utils diff --help` -->
+The single subcommand, `dvc-utils diff`, is also exposed directly as `dvc-dff`:
+
+### `dvc-diff` <a id="dvc-diff"></a>
+<!-- `bmdf -- dvc-diff --help` -->
 ```bash
-dvc-utils diff --help
-# Usage: dvc-utils diff [OPTIONS] [cmd...] <path>
+dvc-diff --help
+# Usage: dvc-diff [OPTIONS] [exec_cmd...] <path>
 #
 #   Diff a file at two commits (or one commit vs. current worktree), optionally
 #   passing both through `cmd` first
@@ -69,6 +71,8 @@ dvc-utils diff --help
 #                            to `diff`)
 #   -v, --verbose            Log intermediate commands to stderr
 #   -w, --ignore-whitespace  Ignore whitespace differences (pass `-w` to `diff`)
+#   -x, --exec-cmd TEXT      Command(s) to execute before diffing; alternate
+#                            syntax to passing commands as positional arguments
 #   --help                   Show this message and exit.
 ```
 
@@ -91,7 +95,7 @@ parquet_schema() {
     parquet2json "$1" schema
 }
 export -f parquet_schema
-dvc-utils diff -r $commit^..$commit parquet_schema $path
+dvc-diff -r $commit^..$commit parquet_schema $path
 ```
 <details><summary>Output</summary>
 
@@ -149,7 +153,7 @@ pretty_print_first_row() {
     parquet2json "$1" cat -l 1 | jq .
 }
 export -f pretty_print_first_row
-dvc-utils diff -r $commit^..$commit pretty_print_first_row $path
+dvc-diff -r $commit^..$commit pretty_print_first_row $path
 ```
 
 <details><summary>Output</summary>
@@ -208,7 +212,7 @@ parquet_row_count() {
     parquet2json "$1" rowcount
 }
 export -f parquet_row_count
-dvc-utils diff -r $commit^..$commit parquet_row_count $path
+dvc-diff -r $commit^..$commit parquet_row_count $path
 ```
 
 This time we get no output; [the given `$commit`][commit] didn't change the row count in the DVC-tracked Parquet file [`$path`][commit path].

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/README.md

@@ -4,7 +4,7 @@ CLI for diffing [DVC] files, optionally passing both through another command fir
 <!-- toc -->
 - [Installation](#installation)
 - [Usage](#usage)
-    - [`dvc-utils diff`](#dvc-utils-diff)
+    - [`dvc-diff`](#dvc-diff)
 - [Examples](#examples)
     - [Parquet file](#parquet-diff)
         - [Schema diff](#parquet-schema-diff)
@@ -31,11 +31,13 @@ dvc-utils --help
 #         worktree), optionally passing both through another command first
 ```
 
-### `dvc-utils diff` <a id="dvc-utils-diff"></a>
-<!-- `bmdf -- dvc-utils diff --help` -->
+The single subcommand, `dvc-utils diff`, is also exposed directly as `dvc-dff`:
+
+### `dvc-diff` <a id="dvc-diff"></a>
+<!-- `bmdf -- dvc-diff --help` -->
 ```bash
-dvc-utils diff --help
-# Usage: dvc-utils diff [OPTIONS] [cmd...] <path>
+dvc-diff --help
+# Usage: dvc-diff [OPTIONS] [exec_cmd...] <path>
 #
 #   Diff a file at two commits (or one commit vs. current worktree), optionally
 #   passing both through `cmd` first
@@ -58,6 +60,8 @@ dvc-utils diff --help
 #                            to `diff`)
 #   -v, --verbose            Log intermediate commands to stderr
 #   -w, --ignore-whitespace  Ignore whitespace differences (pass `-w` to `diff`)
+#   -x, --exec-cmd TEXT      Command(s) to execute before diffing; alternate
+#                            syntax to passing commands as positional arguments
 #   --help                   Show this message and exit.
 ```
 
@@ -80,7 +84,7 @@ parquet_schema() {
     parquet2json "$1" schema
 }
 export -f parquet_schema
-dvc-utils diff -r $commit^..$commit parquet_schema $path
+dvc-diff -r $commit^..$commit parquet_schema $path
 ```
 <details><summary>Output</summary>
 
@@ -138,7 +142,7 @@ pretty_print_first_row() {
     parquet2json "$1" cat -l 1 | jq .
 }
 export -f pretty_print_first_row
-dvc-utils diff -r $commit^..$commit pretty_print_first_row $path
+dvc-diff -r $commit^..$commit pretty_print_first_row $path
 ```
 
 <details><summary>Output</summary>
@@ -197,7 +201,7 @@ parquet_row_count() {
     parquet2json "$1" rowcount
 }
 export -f parquet_row_count
-dvc-utils diff -r $commit^..$commit parquet_row_count $path
+dvc-diff -r $commit^..$commit parquet_row_count $path
 ```
 
 This time we get no output; [the given `$commit`][commit] didn't change the row count in the DVC-tracked Parquet file [`$path`][commit path].

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/dvc_utils/main.py

@@ -2,15 +2,12 @@ from functools import cache
 from os import environ as env, getcwd
 from os.path import join, relpath
 import shlex
-from subprocess import Popen, PIPE
 from typing import Optional, Tuple
 
 from click import option, argument, group
 import click
 import yaml
-from utz import process, singleton, err
-
-from dvc_utils.named_pipes import named_pipes
+from utz import diff_cmds, process, err, singleton
 
 
 @group()
@@ -69,86 +66,6 @@ def dvc_cache_path(ref: str, dvc_path: Optional[str] = None, log: bool = False)
     return join(dvc_cache_dir(log=log), 'files', 'md5', dirname, basename)
 
 
-def diff_cmds(
-    cmds1: list[str],
-    cmds2: list[str],
-    verbose: bool = False,
-    color: bool = False,
-    unified: int | None = None,
-    ignore_whitespace: bool = False,
-    **kwargs,
-):
-    """Run two sequences of piped commands and diff their output.
-
-    Args:
-        cmds1: First sequence of commands to pipe together
-        cmds2: Second sequence of commands to pipe together
-        verbose: Whether to print commands being executed
-        color: Whether to show colored diff output
-        unified: Number of unified context lines, or None
-        ignore_whitespace: Whether to ignore whitespace changes
-        **kwargs: Additional arguments passed to subprocess.Popen
-
-    Each command sequence will be piped together before being compared.
-    For example, if cmds1 = ['cat foo.txt', 'sort'], the function will
-    execute 'cat foo.txt | sort' before comparing with cmds2's output.
-
-    Adapted from https://stackoverflow.com/a/28840955"""
-    with named_pipes(n=2) as pipes:
-        (pipe1, pipe2) = pipes
-        diff_cmd = [
-            'diff',
-            *(['-w'] if ignore_whitespace else []),
-            *(['-U', str(unified)] if unified is not None else []),
-            *(['--color=always'] if color else []),
-            pipe1,
-            pipe2,
-        ]
-        diff = Popen(diff_cmd)
-        processes = []
-
-        for pipe, cmds in ((pipe1, cmds1), (pipe2, cmds2)):
-            if verbose:
-                err(f"Running pipeline: {' | '.join(cmds)}")
-
-            # Create the pipeline of processes
-            prev_process = None
-            for i, cmd in enumerate(cmds):
-                is_last = i + 1 == len(cmds)
-
-                # For the first process, take input from the original source
-                stdin = None if prev_process is None else prev_process.stdout
-
-                # For the last process, output to the named pipe
-                if is_last:
-                    with open(pipe, 'wb', 0) as pipe_fd:
-                        proc = Popen(
-                            cmd,
-                            stdin=stdin,
-                            stdout=pipe_fd,
-                            close_fds=True,
-                            **kwargs
-                        )
-                # For intermediate processes, output to a pipe
-                else:
-                    proc = Popen(
-                        cmd,
-                        stdin=stdin,
-                        stdout=PIPE,
-                        close_fds=True,
-                        **kwargs
-                    )
-
-                if prev_process is not None:
-                    prev_process.stdout.close()
-
-                processes.append(proc)
-                prev_process = proc
-
-        for p in [diff] + processes:
-            p.wait()
-
-
 @cli.command('diff', short_help='Diff a DVC-tracked file at two commits (or one commit vs. current worktree), optionally passing both through another command first')
 @option('-c', '--color', is_flag=True, help='Colorize the output')
 @option('-r', '--refspec', default='HEAD', help='<commit 1>..<commit 2> (compare two commits) or <commit> (compare <commit> to the worktree)')

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/dvc_utils.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dvc-utils
-Version: 0.0.5
+Version: 0.0.7
 Summary: CLI for diffing DVC files at two commits (or one commit vs. current worktree), optionally passing both through another command first
 Home-page: https://github.com/runsascoded/dvc-utils
 Author: Ryan Williams
@@ -15,7 +15,7 @@ CLI for diffing [DVC] files, optionally passing both through another command fir
 <!-- toc -->
 - [Installation](#installation)
 - [Usage](#usage)
-    - [`dvc-utils diff`](#dvc-utils-diff)
+    - [`dvc-diff`](#dvc-diff)
 - [Examples](#examples)
     - [Parquet file](#parquet-diff)
         - [Schema diff](#parquet-schema-diff)
@@ -42,11 +42,13 @@ dvc-utils --help
 #         worktree), optionally passing both through another command first
 ```
 
-### `dvc-utils diff` <a id="dvc-utils-diff"></a>
-<!-- `bmdf -- dvc-utils diff --help` -->
+The single subcommand, `dvc-utils diff`, is also exposed directly as `dvc-dff`:
+
+### `dvc-diff` <a id="dvc-diff"></a>
+<!-- `bmdf -- dvc-diff --help` -->
 ```bash
-dvc-utils diff --help
-# Usage: dvc-utils diff [OPTIONS] [cmd...] <path>
+dvc-diff --help
+# Usage: dvc-diff [OPTIONS] [exec_cmd...] <path>
 #
 #   Diff a file at two commits (or one commit vs. current worktree), optionally
 #   passing both through `cmd` first
@@ -69,6 +71,8 @@ dvc-utils diff --help
 #                            to `diff`)
 #   -v, --verbose            Log intermediate commands to stderr
 #   -w, --ignore-whitespace  Ignore whitespace differences (pass `-w` to `diff`)
+#   -x, --exec-cmd TEXT      Command(s) to execute before diffing; alternate
+#                            syntax to passing commands as positional arguments
 #   --help                   Show this message and exit.
 ```
 
@@ -91,7 +95,7 @@ parquet_schema() {
     parquet2json "$1" schema
 }
 export -f parquet_schema
-dvc-utils diff -r $commit^..$commit parquet_schema $path
+dvc-diff -r $commit^..$commit parquet_schema $path
 ```
 <details><summary>Output</summary>
 
@@ -149,7 +153,7 @@ pretty_print_first_row() {
     parquet2json "$1" cat -l 1 | jq .
 }
 export -f pretty_print_first_row
-dvc-utils diff -r $commit^..$commit pretty_print_first_row $path
+dvc-diff -r $commit^..$commit pretty_print_first_row $path
 ```
 
 <details><summary>Output</summary>
@@ -208,7 +212,7 @@ parquet_row_count() {
     parquet2json "$1" rowcount
 }
 export -f parquet_row_count
-dvc-utils diff -r $commit^..$commit parquet_row_count $path
+dvc-diff -r $commit^..$commit parquet_row_count $path
 ```
 
 This time we get no output; [the given `$commit`][commit] didn't change the row count in the DVC-tracked Parquet file [`$path`][commit path].

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/dvc_utils.egg-info/SOURCES.txt

@@ -3,7 +3,6 @@ README.md
 setup.py
 dvc_utils/__init__.py
 dvc_utils/main.py
-dvc_utils/named_pipes.py
 dvc_utils.egg-info/PKG-INFO
 dvc_utils.egg-info/SOURCES.txt
 dvc_utils.egg-info/dependency_links.txt

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/dvc_utils.egg-info/entry_points.txt

@@ -1,2 +1,3 @@
 [console_scripts]
+dvc-diff = dvc_utils.main:dvc_utils_diff
 dvc-utils = dvc_utils.main:cli

{dvc-utils-0.0.5 → dvc-utils-0.0.7}/setup.py

@@ -2,7 +2,7 @@ from setuptools import setup
 
 setup(
     name='dvc-utils',
-    version="0.0.5",
+    version="0.0.7",
     description="CLI for diffing DVC files at two commits (or one commit vs. current worktree), optionally passing both through another command first",
     long_description=open("README.md").read(),
     long_description_content_type="text/markdown",
@@ -10,6 +10,7 @@ setup(
     entry_points={
         'console_scripts': [
             'dvc-utils = dvc_utils.main:cli',
+            'dvc-diff = dvc_utils.main:dvc_utils_diff',
         ],
     },
     license="MIT",

dvc-utils-0.0.5/dvc_utils/named_pipes.py

@@ -1,19 +0,0 @@
-import os
-import shutil
-import tempfile
-from contextlib import contextmanager
-
-
-@contextmanager
-def named_pipes(n: int = 1):
-    """Yield a list of paths to named pipes that are created and destroyed
-
-    From https://stackoverflow.com/a/28840955"""
-    dirname = tempfile.mkdtemp()
-    try:
-        paths = [os.path.join(dirname, 'named_pipe' + str(i)) for i in range(n)]
-        for path in paths:
-            os.mkfifo(path)
-        yield paths
-    finally:
-        shutil.rmtree(dirname)