mirror of
https://github.com/Stone-Red-Code/dvc.git
synced 2026-11-02 07:18:25 +01:00
metrics: rework metrics implementation (#3658)
Dropping support for all types except json and yaml. No more xpath support. Fixes #3572
This commit is contained in:
Ruslan Kuprieiev
+32
-89
@@ -14,11 +14,9 @@ logger = logging.getLogger(__name__)
|
||||
def show_metrics(
|
||||
metrics, all_branches=False, all_tags=False, all_commits=False
|
||||
):
|
||||
"""
|
||||
Args:
|
||||
metrics (list): Where each element is either a `list`
|
||||
if an xpath was specified, otherwise a `str`
|
||||
"""
|
||||
from flatten_json import flatten
|
||||
from dvc.utils.diff import format_dict
|
||||
|
||||
# When `metrics` contains a `None` key, it means that some files
|
||||
# specified as `targets` in `repo.metrics.show` didn't contain any metrics.
|
||||
missing = metrics.pop(None, None)
|
||||
@@ -28,21 +26,13 @@ def show_metrics(
|
||||
logger.info("{branch}:".format(branch=branch))
|
||||
|
||||
for fname, metric in val.items():
|
||||
if isinstance(metric, dict):
|
||||
lines = list(metric.values())
|
||||
elif isinstance(metric, list):
|
||||
lines = metric
|
||||
else:
|
||||
lines = metric.splitlines()
|
||||
if not isinstance(metric, dict):
|
||||
logger.info("\t{}: {}".format(fname, str(metric)))
|
||||
continue
|
||||
|
||||
if len(lines) > 1:
|
||||
logger.info("\t{fname}:".format(fname=fname))
|
||||
|
||||
for line in lines:
|
||||
logger.info("\t\t{content}".format(content=line))
|
||||
|
||||
else:
|
||||
logger.info("\t{}: {}".format(fname, metric))
|
||||
logger.info("\t{}:".format(fname))
|
||||
for key, value in flatten(format_dict(metric), ".").items():
|
||||
logger.info("\t\t{}: {}".format(key, value))
|
||||
|
||||
if missing:
|
||||
raise BadMetricError(missing)
|
||||
@@ -53,14 +43,17 @@ class CmdMetricsShow(CmdBase):
|
||||
try:
|
||||
metrics = self.repo.metrics.show(
|
||||
self.args.targets,
|
||||
typ=self.args.type,
|
||||
xpath=self.args.xpath,
|
||||
all_branches=self.args.all_branches,
|
||||
all_tags=self.args.all_tags,
|
||||
all_commits=self.args.all_commits,
|
||||
recursive=self.args.recursive,
|
||||
)
|
||||
|
||||
if self.args.show_json:
|
||||
import json
|
||||
|
||||
logger.info(json.dumps(metrics))
|
||||
else:
|
||||
show_metrics(
|
||||
metrics,
|
||||
self.args.all_branches,
|
||||
@@ -74,25 +67,10 @@ class CmdMetricsShow(CmdBase):
|
||||
return 0
|
||||
|
||||
|
||||
class CmdMetricsModify(CmdBase):
|
||||
def run(self):
|
||||
try:
|
||||
self.repo.metrics.modify(
|
||||
self.args.path, typ=self.args.type, xpath=self.args.xpath
|
||||
)
|
||||
except DvcException:
|
||||
logger.exception("failed to modify metric file settings")
|
||||
return 1
|
||||
|
||||
return 0
|
||||
|
||||
|
||||
class CmdMetricsAdd(CmdBase):
|
||||
def run(self):
|
||||
try:
|
||||
self.repo.metrics.add(
|
||||
self.args.path, self.args.type, self.args.xpath
|
||||
)
|
||||
self.repo.metrics.add(self.args.path)
|
||||
except DvcException:
|
||||
msg = "failed to add metric file '{}'".format(self.args.path)
|
||||
logger.exception(msg)
|
||||
@@ -114,11 +92,14 @@ class CmdMetricsRemove(CmdBase):
|
||||
|
||||
|
||||
def _show_diff(diff):
|
||||
from collections import OrderedDict
|
||||
|
||||
from dvc.utils.diff import table
|
||||
|
||||
rows = []
|
||||
for fname, mdiff in diff.items():
|
||||
for metric, change in mdiff.items():
|
||||
sorted_mdiff = OrderedDict(sorted(mdiff.items()))
|
||||
for metric, change in sorted_mdiff.items():
|
||||
rows.append(
|
||||
[
|
||||
fname,
|
||||
@@ -138,9 +119,8 @@ class CmdMetricsDiff(CmdBase):
|
||||
a_rev=self.args.a_rev,
|
||||
b_rev=self.args.b_rev,
|
||||
targets=self.args.targets,
|
||||
typ=self.args.type,
|
||||
xpath=self.args.xpath,
|
||||
recursive=self.args.recursive,
|
||||
all=self.args.all,
|
||||
)
|
||||
|
||||
if self.args.show_json:
|
||||
@@ -185,32 +165,9 @@ def add_parser(subparsers, parent_parser):
|
||||
help=METRICS_ADD_HELP,
|
||||
formatter_class=argparse.RawDescriptionHelpFormatter,
|
||||
)
|
||||
metrics_add_parser.add_argument(
|
||||
"-t", "--type", help="Type of metrics (json/yaml).", metavar="<type>",
|
||||
)
|
||||
metrics_add_parser.add_argument(
|
||||
"-x", "--xpath", help="json/yaml path.", metavar="<path>",
|
||||
)
|
||||
metrics_add_parser.add_argument("path", help="Path to a metric file.")
|
||||
metrics_add_parser.set_defaults(func=CmdMetricsAdd)
|
||||
|
||||
METRICS_MODIFY_HELP = "Modify metric default formatting."
|
||||
metrics_modify_parser = metrics_subparsers.add_parser(
|
||||
"modify",
|
||||
parents=[parent_parser],
|
||||
description=append_doc_link(METRICS_MODIFY_HELP, "metrics/modify"),
|
||||
help=METRICS_MODIFY_HELP,
|
||||
formatter_class=argparse.RawDescriptionHelpFormatter,
|
||||
)
|
||||
metrics_modify_parser.add_argument(
|
||||
"-t", "--type", help="Type of metrics (json/yaml).", metavar="<type>",
|
||||
)
|
||||
metrics_modify_parser.add_argument(
|
||||
"-x", "--xpath", help="json/yaml path.", metavar="<path>",
|
||||
)
|
||||
metrics_modify_parser.add_argument("path", help="Path to a metric file.")
|
||||
metrics_modify_parser.set_defaults(func=CmdMetricsModify)
|
||||
|
||||
METRICS_SHOW_HELP = "Print metrics, with optional formatting."
|
||||
metrics_show_parser = metrics_subparsers.add_parser(
|
||||
"show",
|
||||
@@ -224,19 +181,6 @@ def add_parser(subparsers, parent_parser):
|
||||
nargs="*",
|
||||
help="Metric files or directories (see -R) to show",
|
||||
)
|
||||
metrics_show_parser.add_argument(
|
||||
"-t",
|
||||
"--type",
|
||||
help=(
|
||||
"Type of metrics (json/yaml). "
|
||||
"It can be detected by the file extension automatically. "
|
||||
"Unsupported types will be treated as raw."
|
||||
),
|
||||
metavar="<type>",
|
||||
)
|
||||
metrics_show_parser.add_argument(
|
||||
"-x", "--xpath", help="json/yaml path.", metavar="<path>",
|
||||
)
|
||||
metrics_show_parser.add_argument(
|
||||
"-a",
|
||||
"--all-branches",
|
||||
@@ -267,6 +211,12 @@ def add_parser(subparsers, parent_parser):
|
||||
"metric files."
|
||||
),
|
||||
)
|
||||
metrics_show_parser.add_argument(
|
||||
"--show-json",
|
||||
action="store_true",
|
||||
default=False,
|
||||
help="Show output in JSON format.",
|
||||
)
|
||||
metrics_show_parser.set_defaults(func=CmdMetricsShow)
|
||||
|
||||
METRICS_DIFF_HELP = "Show changes in metrics between commits"
|
||||
@@ -295,19 +245,6 @@ def add_parser(subparsers, parent_parser):
|
||||
),
|
||||
metavar="<paths>",
|
||||
)
|
||||
metrics_diff_parser.add_argument(
|
||||
"-t",
|
||||
"--type",
|
||||
help=(
|
||||
"Type of metrics (json/yaml). "
|
||||
"It can be detected by the file extension automatically. "
|
||||
"Unsupported types will be treated as raw."
|
||||
),
|
||||
metavar="<type>",
|
||||
)
|
||||
metrics_diff_parser.add_argument(
|
||||
"-x", "--xpath", help="json/yaml path.", metavar="<path>",
|
||||
)
|
||||
metrics_diff_parser.add_argument(
|
||||
"-R",
|
||||
"--recursive",
|
||||
@@ -318,6 +255,12 @@ def add_parser(subparsers, parent_parser):
|
||||
"metric files."
|
||||
),
|
||||
)
|
||||
metrics_diff_parser.add_argument(
|
||||
"--all",
|
||||
action="store_true",
|
||||
default=False,
|
||||
help="Show unchanged metrics as well.",
|
||||
)
|
||||
metrics_diff_parser.add_argument(
|
||||
"--show-json",
|
||||
action="store_true",
|
||||
|
||||
@@ -87,6 +87,7 @@ class RemoteLOCAL(RemoteBASE):
|
||||
return self.checksum_to_path_info(md5).url
|
||||
|
||||
def exists(self, path_info):
|
||||
assert is_working_tree(self.repo.tree)
|
||||
assert path_info.scheme == "local"
|
||||
return self.repo.tree.exists(fspath_py35(path_info))
|
||||
|
||||
|
||||
@@ -7,11 +7,6 @@ class Metrics(object):
|
||||
|
||||
return add(self.repo, *args, **kwargs)
|
||||
|
||||
def modify(self, *args, **kwargs):
|
||||
from dvc.repo.metrics.modify import modify
|
||||
|
||||
return modify(self.repo, *args, **kwargs)
|
||||
|
||||
def show(self, *args, **kwargs):
|
||||
from dvc.repo.metrics.show import show
|
||||
|
||||
|
||||
@@ -1,7 +1,5 @@
|
||||
from dvc.repo.metrics.modify import modify
|
||||
|
||||
|
||||
def add(repo, path, typ=None, xpath=None):
|
||||
if not typ:
|
||||
typ = "raw"
|
||||
modify(repo, path, typ, xpath)
|
||||
def add(repo, path):
|
||||
modify(repo, path)
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
import dvc.utils.diff
|
||||
from dvc.utils.diff import format_dict, diff as _diff
|
||||
from dvc.exceptions import NoMetricsError
|
||||
|
||||
|
||||
@@ -13,7 +13,11 @@ def _get_metrics(repo, *args, rev=None, **kwargs):
|
||||
|
||||
|
||||
def diff(repo, *args, a_rev=None, b_rev=None, **kwargs):
|
||||
with_unchanged = kwargs.pop("all", False)
|
||||
|
||||
old = _get_metrics(repo, *args, **kwargs, rev=(a_rev or "HEAD"))
|
||||
new = _get_metrics(repo, *args, **kwargs, rev=b_rev)
|
||||
|
||||
return dvc.utils.diff.diff(old, new)
|
||||
return _diff(
|
||||
format_dict(old), format_dict(new), with_unchanged=with_unchanged
|
||||
)
|
||||
|
||||
@@ -4,8 +4,7 @@ from dvc.repo import locked
|
||||
|
||||
|
||||
@locked
|
||||
def modify(repo, path, typ=None, xpath=None, delete=False):
|
||||
supported_types = ["raw", "json", "csv", "tsv", "hcsv", "htsv"]
|
||||
def modify(repo, path, delete=False):
|
||||
outs = repo.find_outs_by_path(path)
|
||||
assert len(outs) == 1
|
||||
out = outs[0]
|
||||
@@ -14,25 +13,6 @@ def modify(repo, path, typ=None, xpath=None, delete=False):
|
||||
msg = "output '{}' scheme '{}' is not supported for metrics"
|
||||
raise DvcException(msg.format(out.path, out.path_info.scheme))
|
||||
|
||||
if typ is not None:
|
||||
typ = typ.lower().strip()
|
||||
if typ not in ["raw", "json", "csv", "tsv", "hcsv", "htsv"]:
|
||||
msg = (
|
||||
"metric type '{typ}' is not supported, "
|
||||
"must be one of [{types}]"
|
||||
)
|
||||
raise DvcException(
|
||||
msg.format(typ=typ, types=", ".join(supported_types))
|
||||
)
|
||||
if not isinstance(out.metric, dict):
|
||||
out.metric = {}
|
||||
out.metric[out.PARAM_METRIC_TYPE] = typ
|
||||
|
||||
if xpath is not None:
|
||||
if not isinstance(out.metric, dict):
|
||||
out.metric = {}
|
||||
out.metric[out.PARAM_METRIC_XPATH] = xpath
|
||||
|
||||
if delete:
|
||||
out.metric = None
|
||||
|
||||
|
||||
+60
-260
@@ -1,258 +1,79 @@
|
||||
import csv
|
||||
import errno
|
||||
import json
|
||||
import yaml
|
||||
import logging
|
||||
import os
|
||||
import io
|
||||
|
||||
from jsonpath_ng.ext import parse
|
||||
|
||||
from dvc.path_info import PathInfo
|
||||
from dvc.compat import fspath_py35
|
||||
from dvc.exceptions import NoMetricsError
|
||||
from dvc.exceptions import OutputNotFoundError
|
||||
from dvc.repo import locked
|
||||
from dvc.repo.tree import RepoTree
|
||||
|
||||
NO_METRICS_FILE_AT_REFERENCE_WARNING = (
|
||||
"Metrics file '{}' does not exist at the reference '{}'."
|
||||
)
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
def _read_metric_json(fd, json_path):
|
||||
parser = parse(json_path)
|
||||
return {str(x.full_path): x.value for x in parser.find(json.load(fd))}
|
||||
def _collect_metrics(repo, targets, recursive):
|
||||
metrics = set()
|
||||
|
||||
for stage in repo.stages:
|
||||
for out in stage.outs:
|
||||
if not out.metric:
|
||||
continue
|
||||
|
||||
metrics.add(out.path_info)
|
||||
|
||||
if not targets:
|
||||
return list(metrics)
|
||||
|
||||
target_infos = [PathInfo(os.path.abspath(target)) for target in targets]
|
||||
|
||||
def _filter(path_info):
|
||||
for info in target_infos:
|
||||
func = path_info.isin_or_eq if recursive else path_info.__eq__
|
||||
if func(info):
|
||||
return True
|
||||
return False
|
||||
|
||||
return list(filter(_filter, metrics))
|
||||
|
||||
|
||||
def _get_values(row):
|
||||
if isinstance(row, dict):
|
||||
return list(row.values())
|
||||
else:
|
||||
return row
|
||||
def _extract_metrics(metrics):
|
||||
if isinstance(metrics, (int, float)):
|
||||
return metrics
|
||||
|
||||
if not isinstance(metrics, dict):
|
||||
return None
|
||||
|
||||
def _do_read_metric_xsv(reader, row, col):
|
||||
if col is not None and row is not None:
|
||||
return [reader[row][col]]
|
||||
elif col is not None:
|
||||
return [r[col] for r in reader]
|
||||
elif row is not None:
|
||||
return _get_values(reader[row])
|
||||
return [_get_values(r) for r in reader]
|
||||
ret = {}
|
||||
for key, val in metrics.items():
|
||||
m = _extract_metrics(val)
|
||||
if m:
|
||||
ret[key] = m
|
||||
|
||||
|
||||
def _read_metric_hxsv(fd, hxsv_path, delimiter):
|
||||
indices = hxsv_path.split(",")
|
||||
row = indices[0]
|
||||
row = int(row) if row else None
|
||||
col = indices[1] if len(indices) > 1 and indices[1] else None
|
||||
reader = list(csv.DictReader(fd, delimiter=delimiter))
|
||||
return _do_read_metric_xsv(reader, row, col)
|
||||
|
||||
|
||||
def _read_metric_xsv(fd, xsv_path, delimiter):
|
||||
indices = xsv_path.split(",")
|
||||
row = indices[0]
|
||||
row = int(row) if row else None
|
||||
col = int(indices[1]) if len(indices) > 1 and indices[1] else None
|
||||
reader = list(csv.reader(fd, delimiter=delimiter))
|
||||
return _do_read_metric_xsv(reader, row, col)
|
||||
|
||||
|
||||
def _read_typed_metric(typ, xpath, fd):
|
||||
if typ == "json":
|
||||
ret = _read_metric_json(fd, xpath)
|
||||
elif typ == "csv":
|
||||
ret = _read_metric_xsv(fd, xpath, ",")
|
||||
elif typ == "tsv":
|
||||
ret = _read_metric_xsv(fd, xpath, "\t")
|
||||
elif typ == "hcsv":
|
||||
ret = _read_metric_hxsv(fd, xpath, ",")
|
||||
elif typ == "htsv":
|
||||
ret = _read_metric_hxsv(fd, xpath, "\t")
|
||||
else:
|
||||
ret = fd.read().strip()
|
||||
return ret
|
||||
|
||||
|
||||
def _format_csv(content, delimiter):
|
||||
"""Format delimited text to have same column width.
|
||||
def _read_metrics(repo, metrics, rev):
|
||||
tree = RepoTree(repo)
|
||||
|
||||
Args:
|
||||
content (str): The content of a metric.
|
||||
delimiter (str): Value separator
|
||||
|
||||
Returns:
|
||||
str: Formatted content.
|
||||
|
||||
Example:
|
||||
|
||||
>>> content = (
|
||||
"value_mse,deviation_mse,data_set\n"
|
||||
"0.421601,0.173461,train\n"
|
||||
"0.67528,0.289545,testing\n"
|
||||
"0.671502,0.297848,validation\n"
|
||||
)
|
||||
>>> _format_csv(content, ",")
|
||||
|
||||
"value_mse deviation_mse data_set\n"
|
||||
"0.421601 0.173461 train\n"
|
||||
"0.67528 0.289545 testing\n"
|
||||
"0.671502 0.297848 validation\n"
|
||||
"""
|
||||
reader = csv.reader(io.StringIO(content), delimiter=delimiter)
|
||||
rows = [row for row in reader]
|
||||
max_widths = [max(map(len, column)) for column in zip(*rows)]
|
||||
|
||||
lines = [
|
||||
" ".join(
|
||||
"{entry:{width}}".format(entry=entry, width=width + 2)
|
||||
for entry, width in zip(row, max_widths)
|
||||
)
|
||||
for row in rows
|
||||
]
|
||||
|
||||
return "\n".join(lines)
|
||||
|
||||
|
||||
def _format_output(content, typ):
|
||||
"""Tabularize the content according to its type.
|
||||
|
||||
Args:
|
||||
content (str): The content of a metric.
|
||||
typ (str): The type of metric -- (raw|json|tsv|htsv|csv|hcsv).
|
||||
|
||||
Returns:
|
||||
str: Content in a raw or tabular format.
|
||||
"""
|
||||
|
||||
if "csv" in typ:
|
||||
return _format_csv(content, delimiter=",")
|
||||
|
||||
if "tsv" in typ:
|
||||
return _format_csv(content, delimiter="\t")
|
||||
|
||||
return content
|
||||
|
||||
|
||||
def _read_metric(fd, typ=None, xpath=None, fname=None, branch=None):
|
||||
typ = typ.lower().strip() if typ else typ
|
||||
try:
|
||||
if xpath:
|
||||
return _read_typed_metric(typ, xpath.strip(), fd)
|
||||
else:
|
||||
return _format_output(fd.read().strip(), typ)
|
||||
# Json path library has to be replaced or wrapped in
|
||||
# order to fix this too broad except clause.
|
||||
except Exception:
|
||||
logger.exception(
|
||||
"unable to read metric in '{}' in branch '{}'".format(
|
||||
fname, branch
|
||||
)
|
||||
)
|
||||
return None
|
||||
|
||||
|
||||
def _collect_metrics(repo, path, recursive, typ, xpath, branch):
|
||||
"""Gather all the metric outputs.
|
||||
|
||||
Args:
|
||||
path (str): Path to a metric file or a directory.
|
||||
recursive (bool): If path is a directory, do a recursive search for
|
||||
metrics on the given path.
|
||||
typ (str): The type that will be used to interpret the metric file,
|
||||
one of the followings - (raw|json|tsv|htsv|csv|hcsv).
|
||||
xpath (str): Path to search for.
|
||||
branch (str): Branch to look up for metrics.
|
||||
|
||||
Returns:
|
||||
list(tuple): (output, typ, xpath)
|
||||
- output:
|
||||
- typ:
|
||||
- xpath:
|
||||
"""
|
||||
outs = [out for stage in repo.stages for out in stage.outs]
|
||||
|
||||
if path:
|
||||
try:
|
||||
outs = repo.find_outs_by_path(path, outs=outs, recursive=recursive)
|
||||
except OutputNotFoundError:
|
||||
logger.debug(
|
||||
"DVC-file not for found for '{}' in branch '{}'".format(
|
||||
path, branch
|
||||
)
|
||||
)
|
||||
return []
|
||||
|
||||
res = []
|
||||
for o in outs:
|
||||
if not o.metric:
|
||||
continue
|
||||
|
||||
# NOTE this case assumes that typ has not been provided in CLI call
|
||||
# and one of the following cases:
|
||||
# - stage file contains metric type
|
||||
# - typ will be read from file extension later
|
||||
if not typ and isinstance(o.metric, dict):
|
||||
t = o.metric.get(o.PARAM_METRIC_TYPE, typ)
|
||||
# NOTE user might want to check different xpath, hence xpath first
|
||||
x = xpath or o.metric.get(o.PARAM_METRIC_XPATH)
|
||||
else:
|
||||
t = typ
|
||||
x = xpath
|
||||
|
||||
res.append((o, t, x))
|
||||
|
||||
return res
|
||||
|
||||
|
||||
def _read_metrics(repo, metrics, branch):
|
||||
"""Read the content of each metric file and format it.
|
||||
|
||||
Args:
|
||||
metrics (list): List of metric touples
|
||||
branch (str): Branch to look up for metrics.
|
||||
|
||||
Returns:
|
||||
A dict mapping keys with metrics path name and content.
|
||||
For example:
|
||||
|
||||
{'metric.csv': ("value_mse deviation_mse data_set\n"
|
||||
"0.421601 0.173461 train\n"
|
||||
"0.67528 0.289545 testing\n"
|
||||
"0.671502 0.297848 validation\n")}
|
||||
"""
|
||||
res = {}
|
||||
for out, typ, xpath in metrics:
|
||||
assert out.scheme == "local"
|
||||
if not typ:
|
||||
typ = os.path.splitext(out.fspath.lower())[1].replace(".", "")
|
||||
if out.use_cache:
|
||||
open_fun = open
|
||||
path = repo.cache.local.get(out.checksum)
|
||||
else:
|
||||
open_fun = repo.tree.open
|
||||
path = out.fspath
|
||||
try:
|
||||
|
||||
with open_fun(path) as fd:
|
||||
metric = _read_metric(
|
||||
fd, typ=typ, xpath=xpath, fname=str(out), branch=branch
|
||||
)
|
||||
except IOError as e:
|
||||
if e.errno == errno.ENOENT:
|
||||
logger.warning(
|
||||
NO_METRICS_FILE_AT_REFERENCE_WARNING.format(
|
||||
out.path_info, branch
|
||||
)
|
||||
)
|
||||
metric = None
|
||||
else:
|
||||
raise
|
||||
|
||||
if not metric:
|
||||
for metric in metrics:
|
||||
if not tree.exists(fspath_py35(metric)):
|
||||
continue
|
||||
|
||||
res[str(out)] = metric
|
||||
with tree.open(fspath_py35(metric), "r") as fobj:
|
||||
try:
|
||||
# NOTE this also supports JSON
|
||||
val = yaml.safe_load(fobj)
|
||||
except yaml.YAMLError:
|
||||
logger.debug(
|
||||
"failed to read '%s' on '%s'", metric, rev, exc_info=True
|
||||
)
|
||||
continue
|
||||
|
||||
val = _extract_metrics(val)
|
||||
if val:
|
||||
res[str(metric)] = val
|
||||
|
||||
return res
|
||||
|
||||
@@ -261,8 +82,6 @@ def _read_metrics(repo, metrics, branch):
|
||||
def show(
|
||||
repo,
|
||||
targets=None,
|
||||
typ=None,
|
||||
xpath=None,
|
||||
all_branches=False,
|
||||
all_tags=False,
|
||||
recursive=False,
|
||||
@@ -270,34 +89,20 @@ def show(
|
||||
all_commits=False,
|
||||
):
|
||||
res = {}
|
||||
found = set()
|
||||
|
||||
if not targets:
|
||||
# Iterate once to call `_collect_metrics` on all the stages
|
||||
targets = [None]
|
||||
|
||||
for branch in repo.brancher(
|
||||
for rev in repo.brancher(
|
||||
revs=revs,
|
||||
all_branches=all_branches,
|
||||
all_tags=all_tags,
|
||||
all_commits=all_commits,
|
||||
):
|
||||
metrics = {}
|
||||
metrics = _collect_metrics(repo, targets, recursive)
|
||||
vals = _read_metrics(repo, metrics, rev)
|
||||
|
||||
for target in targets:
|
||||
entries = _collect_metrics(
|
||||
repo, target, recursive, typ, xpath, branch
|
||||
)
|
||||
metric = _read_metrics(repo, entries, branch)
|
||||
if vals:
|
||||
res[rev] = vals
|
||||
|
||||
if metric:
|
||||
found.add(target)
|
||||
metrics.update(metric)
|
||||
|
||||
if metrics:
|
||||
res[branch] = metrics
|
||||
|
||||
if not res and not any(targets):
|
||||
if not res:
|
||||
raise NoMetricsError()
|
||||
|
||||
# Hide working tree metrics if they are the same as in the active branch
|
||||
@@ -309,9 +114,4 @@ def show(
|
||||
if res.get("working tree") == res.get(active_branch):
|
||||
res.pop("working tree", None)
|
||||
|
||||
missing = set(targets) - found
|
||||
|
||||
if missing:
|
||||
res[None] = missing
|
||||
|
||||
return res
|
||||
|
||||
+3
-16
@@ -1,4 +1,4 @@
|
||||
import dvc.utils.diff
|
||||
from dvc.utils.diff import format_dict, diff as _diff
|
||||
from .show import NoParamsError
|
||||
|
||||
|
||||
@@ -10,25 +10,12 @@ def _get_params(repo, *args, rev=None, **kwargs):
|
||||
return {}
|
||||
|
||||
|
||||
def _format(params):
|
||||
ret = {}
|
||||
for key, val in params.items():
|
||||
if isinstance(val, dict):
|
||||
new_val = _format(val)
|
||||
elif isinstance(val, list):
|
||||
new_val = str(val)
|
||||
else:
|
||||
new_val = val
|
||||
ret[key] = new_val
|
||||
return ret
|
||||
|
||||
|
||||
def diff(repo, *args, a_rev=None, b_rev=None, **kwargs):
|
||||
with_unchanged = kwargs.pop("all", False)
|
||||
|
||||
old = _get_params(repo, *args, **kwargs, rev=(a_rev or "HEAD"))
|
||||
new = _get_params(repo, *args, **kwargs, rev=b_rev)
|
||||
|
||||
return dvc.utils.diff.diff(
|
||||
_format(old), _format(new), with_unchanged=with_unchanged
|
||||
return _diff(
|
||||
format_dict(old), format_dict(new), with_unchanged=with_unchanged
|
||||
)
|
||||
|
||||
+28
-4
@@ -1,7 +1,7 @@
|
||||
import os
|
||||
import errno
|
||||
|
||||
from dvc.scm.tree import BaseTree
|
||||
from dvc.scm.tree import BaseTree, WorkingTree
|
||||
from dvc.path_info import PathInfo
|
||||
from dvc.exceptions import OutputNotFoundError
|
||||
|
||||
@@ -32,10 +32,17 @@ class DvcTree(BaseTree):
|
||||
raise IOError(errno.EISDIR)
|
||||
|
||||
out = outs[0]
|
||||
if not out.changed_cache():
|
||||
return open(out.cache_path, mode=mode, encoding=encoding)
|
||||
|
||||
# temporary hack to make cache use WorkingTree and not GitTree, because
|
||||
# cache dir doesn't exist in the latter.
|
||||
saved_tree = self.repo.tree
|
||||
self.repo.tree = WorkingTree(self.repo.root_dir)
|
||||
try:
|
||||
if out.changed_cache():
|
||||
raise FileNotFoundError
|
||||
finally:
|
||||
self.repo.tree = saved_tree
|
||||
|
||||
return open(out.cache_path, mode=mode, encoding=encoding)
|
||||
|
||||
def exists(self, path):
|
||||
try:
|
||||
@@ -115,3 +122,20 @@ class DvcTree(BaseTree):
|
||||
|
||||
def isexec(self, path):
|
||||
return False
|
||||
|
||||
|
||||
class RepoTree(BaseTree):
|
||||
def __init__(self, repo):
|
||||
self.repo = repo
|
||||
self.dvctree = DvcTree(repo)
|
||||
|
||||
def open(self, *args, **kwargs):
|
||||
try:
|
||||
return self.dvctree.open(*args, **kwargs)
|
||||
except FileNotFoundError:
|
||||
pass
|
||||
|
||||
return self.repo.tree.open(*args, **kwargs)
|
||||
|
||||
def exists(self, path):
|
||||
return self.repo.tree.exists(path) or self.dvctree.exists(path)
|
||||
|
||||
+16
-3
@@ -1,8 +1,6 @@
|
||||
import json
|
||||
from collections import defaultdict
|
||||
|
||||
from flatten_json import flatten
|
||||
|
||||
|
||||
def _parse(raw):
|
||||
if raw is None or isinstance(raw, (dict, list, int, float)):
|
||||
@@ -37,7 +35,9 @@ def _flatten(d):
|
||||
return defaultdict(lambda: None)
|
||||
|
||||
if isinstance(d, dict):
|
||||
return defaultdict(lambda: None, flatten(d, "."))
|
||||
from flatten_json import flatten as fltn
|
||||
|
||||
return defaultdict(lambda: None, fltn(d, "."))
|
||||
|
||||
return defaultdict(lambda: "unable to parse")
|
||||
|
||||
@@ -103,3 +103,16 @@ def table(header, rows):
|
||||
t.add_rows([header] + rows)
|
||||
|
||||
return t.draw()
|
||||
|
||||
|
||||
def format_dict(d):
|
||||
ret = {}
|
||||
for key, val in d.items():
|
||||
if isinstance(val, dict):
|
||||
new_val = format_dict(val)
|
||||
elif isinstance(val, list):
|
||||
new_val = str(val)
|
||||
else:
|
||||
new_val = val
|
||||
ret[key] = new_val
|
||||
return ret
|
||||
|
||||
@@ -0,0 +1,135 @@
|
||||
import json
|
||||
import yaml
|
||||
|
||||
|
||||
def test_metrics_diff_simple(tmp_dir, scm, dvc):
|
||||
def _gen(val):
|
||||
tmp_dir.gen({"m.yaml": str(val)})
|
||||
dvc.run(cmd="", metrics=["m.yaml"])
|
||||
dvc.scm.add(["m.yaml.dvc"])
|
||||
dvc.scm.commit(str(val))
|
||||
|
||||
_gen(1)
|
||||
_gen(2)
|
||||
_gen(3)
|
||||
|
||||
expected = {"m.yaml": {"": {"old": 1, "new": 3, "diff": 2}}}
|
||||
|
||||
assert dvc.metrics.diff(a_rev="HEAD~2") == expected
|
||||
|
||||
|
||||
def test_metrics_diff_yaml(tmp_dir, scm, dvc):
|
||||
def _gen(val):
|
||||
metrics = {"a": {"b": {"c": val, "d": 1, "e": str(val)}}}
|
||||
tmp_dir.gen({"m.yaml": yaml.dump(metrics)})
|
||||
dvc.run(cmd="", metrics=["m.yaml"])
|
||||
dvc.scm.add(["m.yaml.dvc"])
|
||||
dvc.scm.commit(str(val))
|
||||
|
||||
_gen(1)
|
||||
_gen(2)
|
||||
_gen(3)
|
||||
|
||||
expected = {"m.yaml": {"a.b.c": {"old": 1, "new": 3, "diff": 2}}}
|
||||
|
||||
assert dvc.metrics.diff(a_rev="HEAD~2") == expected
|
||||
|
||||
|
||||
def test_metrics_diff_json(tmp_dir, scm, dvc):
|
||||
def _gen(val):
|
||||
metrics = {"a": {"b": {"c": val, "d": 1, "e": str(val)}}}
|
||||
tmp_dir.gen({"m.json": json.dumps(metrics)})
|
||||
dvc.run(cmd="", metrics=["m.json"])
|
||||
dvc.scm.add(["m.json.dvc"])
|
||||
dvc.scm.commit(str(val))
|
||||
|
||||
_gen(1)
|
||||
_gen(2)
|
||||
_gen(3)
|
||||
|
||||
expected = {"m.json": {"a.b.c": {"old": 1, "new": 3, "diff": 2}}}
|
||||
assert dvc.metrics.diff(a_rev="HEAD~2") == expected
|
||||
|
||||
|
||||
def test_metrics_diff_json_unchanged(tmp_dir, scm, dvc):
|
||||
def _gen(val):
|
||||
metrics = {"a": {"b": {"c": val, "d": 1, "e": str(val)}}}
|
||||
tmp_dir.gen({"m.json": json.dumps(metrics)})
|
||||
dvc.run(cmd="", metrics=["m.json"])
|
||||
dvc.scm.add(["m.json.dvc"])
|
||||
dvc.scm.commit(str(val))
|
||||
|
||||
_gen(1)
|
||||
_gen(2)
|
||||
_gen(1)
|
||||
|
||||
assert dvc.metrics.diff(a_rev="HEAD~2") == {}
|
||||
|
||||
|
||||
def test_metrics_diff_broken_json(tmp_dir, scm, dvc):
|
||||
metrics = {"a": {"b": {"c": 1, "d": 1, "e": "3"}}}
|
||||
tmp_dir.gen({"m.json": json.dumps(metrics)})
|
||||
dvc.run(cmd="", metrics_no_cache=["m.json"])
|
||||
dvc.scm.add(["m.json.dvc", "m.json"])
|
||||
dvc.scm.commit("add metrics")
|
||||
|
||||
(tmp_dir / "m.json").write_text(json.dumps(metrics) + "ma\nlformed\n")
|
||||
|
||||
assert dvc.metrics.diff() == {
|
||||
"m.json": {
|
||||
"a.b.c": {"old": 1, "new": None},
|
||||
"a.b.d": {"old": 1, "new": None},
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
def test_metrics_diff_no_metrics(tmp_dir, scm, dvc):
|
||||
tmp_dir.scm_gen({"foo": "foo"}, commit="add foo")
|
||||
assert dvc.metrics.diff(a_rev="HEAD~1") == {}
|
||||
|
||||
|
||||
def test_metrics_diff_new_metric(tmp_dir, scm, dvc):
|
||||
metrics = {"a": {"b": {"c": 1, "d": 1, "e": "3"}}}
|
||||
tmp_dir.gen({"m.json": json.dumps(metrics)})
|
||||
dvc.run(cmd="", metrics_no_cache=["m.json"])
|
||||
|
||||
assert dvc.metrics.diff() == {
|
||||
"m.json": {
|
||||
"a.b.c": {"old": None, "new": 1},
|
||||
"a.b.d": {"old": None, "new": 1},
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
def test_metrics_diff_deleted_metric(tmp_dir, scm, dvc):
|
||||
metrics = {"a": {"b": {"c": 1, "d": 1, "e": "3"}}}
|
||||
tmp_dir.gen({"m.json": json.dumps(metrics)})
|
||||
dvc.run(cmd="", metrics_no_cache=["m.json"])
|
||||
dvc.scm.add(["m.json.dvc", "m.json"])
|
||||
dvc.scm.commit("add metrics")
|
||||
|
||||
(tmp_dir / "m.json").unlink()
|
||||
|
||||
assert dvc.metrics.diff() == {
|
||||
"m.json": {
|
||||
"a.b.c": {"old": 1, "new": None},
|
||||
"a.b.d": {"old": 1, "new": None},
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
def test_metrics_diff_with_unchanged(tmp_dir, scm, dvc):
|
||||
tmp_dir.gen("metrics.yaml", "foo: 1\nxyz: 10")
|
||||
dvc.run(metrics_no_cache=["metrics.yaml"])
|
||||
scm.add(["metrics.yaml", "metrics.yaml.dvc"])
|
||||
scm.commit("1")
|
||||
|
||||
tmp_dir.scm_gen("metrics.yaml", "foo: 2\nxyz: 10", commit="2")
|
||||
tmp_dir.scm_gen("metrics.yaml", "foo: 3\nxyz: 10", commit="3")
|
||||
|
||||
assert dvc.metrics.diff(a_rev="HEAD~2", all=True) == {
|
||||
"metrics.yaml": {
|
||||
"foo": {"old": 1, "new": 3, "diff": 2},
|
||||
"xyz": {"old": 10, "new": 10, "diff": 0},
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,50 @@
|
||||
import pytest
|
||||
|
||||
from dvc.repo.metrics.show import NoMetricsError
|
||||
|
||||
|
||||
def test_show_empty(dvc):
|
||||
with pytest.raises(NoMetricsError):
|
||||
dvc.metrics.show()
|
||||
|
||||
|
||||
def test_show_simple(tmp_dir, dvc):
|
||||
tmp_dir.gen("metrics.yaml", "1.1")
|
||||
dvc.run(metrics=["metrics.yaml"])
|
||||
assert dvc.metrics.show() == {"": {"metrics.yaml": 1.1}}
|
||||
|
||||
|
||||
def test_show(tmp_dir, dvc):
|
||||
tmp_dir.gen("metrics.yaml", "foo: 1.1")
|
||||
dvc.run(metrics=["metrics.yaml"])
|
||||
assert dvc.metrics.show() == {"": {"metrics.yaml": {"foo": 1.1}}}
|
||||
|
||||
|
||||
def test_show_multiple(tmp_dir, dvc):
|
||||
tmp_dir.gen("foo", "foo: 1\n")
|
||||
tmp_dir.gen("baz", "baz: 2\n")
|
||||
dvc.run(fname="foo.dvc", metrics=["foo"])
|
||||
dvc.run(fname="baz.dvc", metrics=["baz"])
|
||||
assert dvc.metrics.show() == {"": {"foo": {"foo": 1}, "baz": {"baz": 2}}}
|
||||
|
||||
|
||||
def test_show_invalid_metric(tmp_dir, dvc):
|
||||
tmp_dir.gen("metrics.yaml", "foo:\n- bar\n- baz\nxyz: string")
|
||||
dvc.run(metrics=["metrics.yaml"])
|
||||
with pytest.raises(NoMetricsError):
|
||||
dvc.metrics.show()
|
||||
|
||||
|
||||
def test_show_branch(tmp_dir, scm, dvc):
|
||||
tmp_dir.gen("metrics.yaml", "foo: 1")
|
||||
dvc.run(metrics_no_cache=["metrics.yaml"])
|
||||
scm.add(["metrics.yaml", "metrics.yaml.dvc"])
|
||||
scm.commit("init")
|
||||
|
||||
with tmp_dir.branch("branch", new=True):
|
||||
tmp_dir.scm_gen("metrics.yaml", "foo: 2", commit="branch")
|
||||
|
||||
assert dvc.metrics.show(revs=["branch"]) == {
|
||||
"working tree": {"metrics.yaml": {"foo": 1}},
|
||||
"branch": {"metrics.yaml": {"foo": 2}},
|
||||
}
|
||||
@@ -69,6 +69,7 @@ def test_init_no_scm_cli(tmp_dir):
|
||||
|
||||
|
||||
def test_init_quiet_should_not_display_welcome_screen(tmp_dir, scm, caplog):
|
||||
with caplog.at_level(logging.INFO, logger="dvc"):
|
||||
ret = main(["init", "--quiet"])
|
||||
|
||||
assert 0 == ret
|
||||
|
||||
File diff suppressed because it is too large
Load Diff
@@ -9,11 +9,8 @@ def test_metrics_diff(dvc, mocker):
|
||||
"diff",
|
||||
"HEAD~10",
|
||||
"HEAD~1",
|
||||
"-t",
|
||||
"json",
|
||||
"-x",
|
||||
"x.path",
|
||||
"-R",
|
||||
"--all",
|
||||
"--show-json",
|
||||
"--targets",
|
||||
"target1",
|
||||
@@ -32,9 +29,8 @@ def test_metrics_diff(dvc, mocker):
|
||||
targets=["target1", "target2"],
|
||||
a_rev="HEAD~10",
|
||||
b_rev="HEAD~1",
|
||||
typ="json",
|
||||
xpath="x.path",
|
||||
recursive=True,
|
||||
all=True,
|
||||
)
|
||||
|
||||
|
||||
@@ -90,10 +86,6 @@ def test_metrics_show(dvc, mocker):
|
||||
[
|
||||
"metrics",
|
||||
"show",
|
||||
"-t",
|
||||
"json",
|
||||
"-x",
|
||||
"x.path",
|
||||
"-R",
|
||||
"--all-tags",
|
||||
"--all-branches",
|
||||
@@ -112,8 +104,6 @@ def test_metrics_show(dvc, mocker):
|
||||
m.assert_called_once_with(
|
||||
cmd.repo,
|
||||
["target1", "target2"],
|
||||
typ="json",
|
||||
xpath="x.path",
|
||||
recursive=True,
|
||||
all_tags=True,
|
||||
all_branches=True,
|
||||
@@ -128,3 +118,20 @@ def test_metrics_diff_prec():
|
||||
" Path Metric Value Change\n"
|
||||
"other.json a.b 0.0043 0.0001"
|
||||
)
|
||||
|
||||
|
||||
def test_metrics_diff_sorted():
|
||||
assert _show_diff(
|
||||
{
|
||||
"metrics.yaml": {
|
||||
"x.b": {"old": 5, "new": 6, "diff": 1},
|
||||
"a.d.e": {"old": 3, "new": 4, "diff": 1},
|
||||
"a.b.c": {"old": 1, "new": 2, "diff": 1},
|
||||
}
|
||||
}
|
||||
) == (
|
||||
" Path Metric Value Change\n"
|
||||
"metrics.yaml a.b.c 2 1 \n"
|
||||
"metrics.yaml a.d.e 4 1 \n"
|
||||
"metrics.yaml x.b 6 1 "
|
||||
)
|
||||
|
||||
@@ -24,6 +24,26 @@ def test_open(tmp_dir, dvc):
|
||||
assert fobj.read() == "foo"
|
||||
|
||||
|
||||
def test_open_in_history(tmp_dir, scm, dvc):
|
||||
tmp_dir.gen("foo", "foo")
|
||||
dvc.add("foo")
|
||||
dvc.scm.add(["foo.dvc", ".gitignore"])
|
||||
dvc.scm.commit("foo")
|
||||
|
||||
tmp_dir.gen("foo", "foofoo")
|
||||
dvc.add("foo")
|
||||
dvc.scm.add(["foo.dvc", ".gitignore"])
|
||||
dvc.scm.commit("foofoo")
|
||||
|
||||
for rev in dvc.brancher(revs=["HEAD~1"]):
|
||||
if rev == "working tree":
|
||||
continue
|
||||
|
||||
tree = DvcTree(dvc)
|
||||
with tree.open("foo", "r") as fobj:
|
||||
assert fobj.read() == "foo"
|
||||
|
||||
|
||||
def test_isdir_isfile(tmp_dir, dvc):
|
||||
tmp_dir.gen({"datafile": "data", "datadir": {"foo": "foo", "bar": "bar"}})
|
||||
|
||||
|
||||
Reference in New Issue
Block a user