From 4b77e5d349b735725db3c4d023eaa5fa8a671af4 Mon Sep 17 00:00:00 2001
From: Michael Boniface <m.j.boniface@soton.ac.uk>
Date: Wed, 26 Feb 2025 22:11:32 +0000
Subject: [PATCH] fix: yaml library dump strips all quotes from the output and
there's no configuration. SO you need to create a bespoke Dumper which has
been added to a new module util.py because dumping yaml is done across many
files. closes #41
---
acmc/omop.py | 11 +++++++++--
acmc/phen.py | 29 +++++++++++++++++++++++++----
acmc/trud.py | 11 +++++++++--
acmc/util.py | 7 +++++++
4 files changed, 50 insertions(+), 8 deletions(-)
create mode 100644 acmc/util.py
diff --git a/acmc/omop.py b/acmc/omop.py
index dd5a461..d3f1710 100644
--- a/acmc/omop.py
+++ b/acmc/omop.py
@@ -9,7 +9,7 @@ import json
import yaml
from pathlib import Path
-from acmc import logging_config
+from acmc import util, logging_config
# setup logging
logger = logging_config.setup_logger()
@@ -123,7 +123,14 @@ def write_version_file(version):
"""Writes the OMOP vocaburaries and version to a file"""
vocabularies["version"] = version
with open(VERSION_PATH, "w") as file:
- yaml.dump(vocabularies, file, default_flow_style=False, sort_keys=False)
+ yaml.dump(
+ vocabularies,
+ file,
+ Dumper=util.QuotedDumper,
+ default_flow_style=False,
+ sort_keys=False,
+ default_style='"',
+ )
def clear(db_path):
diff --git a/acmc/phen.py b/acmc/phen.py
index 50b166b..23c8ac7 100644
--- a/acmc/phen.py
+++ b/acmc/phen.py
@@ -17,7 +17,7 @@ from pathlib import Path
from urllib.parse import urlparse, urlunparse
import acmc
-from acmc import trud, omop, parse
+from acmc import trud, omop, parse, util
# setup logging
import acmc.logging_config as lc
@@ -275,7 +275,14 @@ def init(phen_dir, remote_url):
}
with open(phen_path / CONFIG_FILE, "w") as file:
- yaml.dump(config, file, default_flow_style=False, sort_keys=False)
+ yaml.dump(
+ config,
+ file,
+ Dumper=util.QuotedDumper,
+ default_flow_style=False,
+ sort_keys=False,
+ default_style='"',
+ )
# add git ignore
ignore_content = """# Ignore SQLite database files
@@ -611,7 +618,14 @@ def write_vocab_version(phen_path):
}
with open(phen_path / VOCAB_VERSION_FILE, "w") as file:
- yaml.dump(version_data, file, default_flow_style=False, sort_keys=False)
+ yaml.dump(
+ version_data,
+ file,
+ Dumper=util.QuotedDumper,
+ default_flow_style=False,
+ sort_keys=False,
+ default_style='"',
+ )
def map(phen_dir, target_code_type):
@@ -790,7 +804,14 @@ def publish(phen_dir, remote_url):
logger.debug(f"New version: {version}")
config["phenotype"]["version"] = version
with open(config_path, "w") as file:
- yaml.dump(config, file, default_flow_style=False, sort_keys=False)
+ yaml.dump(
+ config,
+ file,
+ Dumper=util.QuotedDumper,
+ default_flow_style=False,
+ sort_keys=False,
+ default_style='"',
+ )
# Add and commit changes to repo
commit_message = f"Committing updates to phenotype {phen_path}"
diff --git a/acmc/trud.py b/acmc/trud.py
index 93298f9..08f5c4a 100644
--- a/acmc/trud.py
+++ b/acmc/trud.py
@@ -11,7 +11,7 @@ import yaml
from pathlib import Path
# setup logging
-import acmc.logging_config as lc
+from acmc import util, logging_config as lc
logger = lc.setup_logger()
@@ -384,7 +384,14 @@ def install():
data = [{k: v for k, v in d.items() if k != "extract"} for d in items]
# save TRUD versions to file to main record of what was downloaded
with open(VERSION_PATH, "w") as file:
- yaml.dump(data, file, default_flow_style=False, sort_keys=False)
+ yaml.dump(
+ data,
+ file,
+ Dumper=util.QuotedDumper,
+ default_flow_style=False,
+ sort_keys=False,
+ default_style='"',
+ )
# Validate and process each item ID
for item in items:
diff --git a/acmc/util.py b/acmc/util.py
new file mode 100644
index 0000000..01bb458
--- /dev/null
+++ b/acmc/util.py
@@ -0,0 +1,7 @@
+import yaml
+
+
+# Custom Dumper to retain quotes on strings in yaml library
+class QuotedDumper(yaml.Dumper):
+ def increase_indent(self, flow=False, indentless=False):
+ return super(QuotedDumper, self).increase_indent(flow, indentless)
--
GitLab