Pyteryx/alteryx_runner/tools/inout/output_data.py

131 lines
5.3 KiB
Python

from __future__ import annotations
import csv
import io
from pathlib import Path
from typing import Dict
import xml.etree.ElementTree as ET
import polars as pl
from tools.base import BaseTool
class OutputDataTool(BaseTool):
def execute(self, inputs: Dict[str, pl.DataFrame]) -> Dict[str, pl.DataFrame]:
df = inputs.get("Input", pl.DataFrame())
if self.config is None or df.is_empty():
return {}
file_el = self.config.find("File")
if file_el is None or not file_el.text:
return {}
raw_path = (file_el.text or "").strip()
fmt = int(file_el.attrib.get("FileFormat", "0"))
max_records_str = (file_el.attrib.get("MaxRecords") or "").strip()
max_records = int(max_records_str) if max_records_str else None
opts = self.config.find("FormatSpecificOptions")
if opts is None:
opts = ET.Element("x")
multi_el = self.config.find("MultiFile")
multi_file = (multi_el.attrib.get("value", "False") if multi_el is not None else "False") == "True"
multi_field = (self.config.findtext("MultiFileField") or "").strip()
multi_type = (self.config.findtext("MultiFileType") or "Suffix").strip()
keep_field = (self.config.findtext("KeepField") or "True").strip().lower() == "true"
out_path = self.ctx.resolve_output_path(raw_path)
out_path.parent.mkdir(parents=True, exist_ok=True)
if multi_file and multi_field and multi_field in df.columns:
for val in df[multi_field].unique().to_list():
part = df.filter(pl.col(multi_field) == val)
if not keep_field:
part = part.drop(multi_field)
part_path = self._multi_path(out_path, str(val), multi_type)
self._write(part, part_path, fmt, opts)
elif max_records:
chunk_num = 0
for i in range(0, len(df), max_records):
chunk = df.slice(i, max_records)
chunk_path = out_path if chunk_num == 0 else out_path.with_stem(
f"{out_path.stem}_{chunk_num}"
)
self._write(chunk, chunk_path, fmt, opts)
chunk_num += 1
else:
self._write(df, out_path, fmt, opts)
if self.ctx.verbose:
print(f"[Output] Wrote {len(df)} rows → {out_path}")
return {}
def _multi_path(self, base: Path, value: str, mode: str) -> Path:
safe = "".join(c if c.isalnum() or c in "-_" else "_" for c in value)
if mode == "Suffix":
return base.with_stem(f"{base.stem}_{safe}")
return base.with_stem(f"{safe}_{base.stem}")
def _write(self, df: pl.DataFrame, path: Path, fmt: int, opts: ET.Element) -> None:
if fmt in (0, 6): # CSV
delim = opts.findtext("Delimeter") or opts.findtext("Delimiter") or ","
# Accept both HeaderRow and Headers attribute names
header_val = (
opts.findtext("HeaderRow")
or opts.findtext("Headers")
or "True"
)
header = header_val.lower() != "false"
line_end = (opts.findtext("LineEndStyle") or "LF").strip().upper()
eol = "\r\n" if line_end == "CRLF" else "\n"
force_quotes = (opts.findtext("ForceQuotes") or "False").lower() == "true"
self._write_csv_alteryx(df, path, delim, header, eol, force_quotes)
elif fmt == 25: # Excel
df.write_excel(str(path))
elif fmt == 2: # Parquet
df.write_parquet(str(path))
elif fmt == 19: # YXDB — fall back to Parquet
fallback = path.with_suffix(".parquet")
df.write_parquet(str(fallback))
if self.ctx.verbose:
print(f"[Output] YXDB write not supported; wrote Parquet to {fallback}")
else:
df.write_csv(str(path))
@staticmethod
def _write_csv_alteryx(
df: pl.DataFrame,
path: Path,
delim: str,
header: bool,
eol: str,
force_quotes: bool,
) -> None:
"""Write CSV matching Alteryx quoting behaviour.
Alteryx quotes a field when it contains the delimiter, a newline,
a double-quote, or a single-quote (apostrophe). We replicate that
by using Python's csv module with a custom quoting function.
"""
# Characters that trigger quoting (beyond csv.QUOTE_MINIMAL's default set)
_needs_quote_chars = frozenset({delim, '"', "'", '\n', '\r'})
def _needs_quoting(val: str) -> bool:
return force_quotes or any(c in _needs_quote_chars for c in val)
with open(path, "w", newline="", encoding="utf-8") as f:
if header:
f.write(delim.join(df.columns) + eol)
for row in df.iter_rows():
parts: list[str] = []
for val in row:
if val is None:
parts.append("")
else:
s = str(val)
if _needs_quoting(s):
# Escape embedded double-quotes by doubling them
s = '"' + s.replace('"', '""') + '"'
parts.append(s)
f.write(delim.join(parts) + eol)