FORMAT_CONVERSION_MAXCOMPUTE/format_convert/convert_xlsx.py

import inspect
import os
import sys
sys.path.append(os.path.dirname(__file__) + "/../")
from format_convert.convert_tree import _Document, _Page, _Table
import logging
import traceback
import pandas
import numpy as np
from format_convert.utils import get_logger, log, memory_decorator
from format_convert.wrapt_timeout_decorator import timeout


@memory_decorator
def xlsx2text(path, unique_type_dir):
    log("into xlsx2text")
    try:
        try:
            # sheet_name=None, 即拿取所有sheet，存为dict
            df_dict = pandas.read_excel(path, header=None, keep_default_na=False, sheet_name=None)
        except Exception as e:
            log("xlsx format error!")
            return [-3]

        df_list = [sheet for sheet in df_dict.values()]
        sheet_text = ""
        for df in df_list:
            text = '<table border="1">' + "\n"
            for index, row in df.iterrows():
                text = text + "<tr>"
                for r in row:
                    text = text + "<td>" + str(r) + "</td>" + "\n"
                    # print(text)
                text = text + "</tr>" + "\n"
            text = text + "</table>" + "\n"
            sheet_text += text

        return [sheet_text]
    except Exception as e:
        log("xlsx2text error!")
        traceback.print_exc()
        return [-1]


class XlsxConvert:

    def __init__(self, path, unique_type_dir):
        self._doc = _Document(path)
        self.path = path
        self.unique_type_dir = unique_type_dir

    @timeout(30, timeout_exception=TimeoutError, use_signals=False)
    def read(self):
        df = pandas.read_excel(self.path, header=None, keep_default_na=False, sheet_name=None)
        return df

    def init_package(self):
        # 各个包初始化
        try:
            self.df = self.read()
            self.sheet_list = [sheet for sheet in self.df.values()]

            # 防止读太多空列空行
            self.col_limit = 100
            self.row_limit = 2000
            self.re_read = 0
            for s in self.sheet_list:
                if s.shape[1] > self.col_limit and s.shape[0] > self.row_limit:
                    self.re_read = 3
                    break
                elif s.shape[0] > self.row_limit:
                    self.re_read = 2
                    break
                elif s.shape[1] > self.col_limit:
                    self.re_read = 1
                    break

            if self.re_read == 3:
                self.df = pandas.read_excel(self.path, header=None, keep_default_na=False,
                                            sheet_name=None, usecols=[x for x in range(self.col_limit)],
                                            nrows=self.row_limit)
            if self.re_read == 2:
                self.df = pandas.read_excel(self.path, header=None, keep_default_na=False,
                                            sheet_name=None, nrows=self.row_limit)
            elif self.re_read == 1:
                self.df = pandas.read_excel(self.path, header=None, keep_default_na=False,
                                            sheet_name=None, usecols=[x for x in range(self.col_limit)])
            if self.re_read > 0:
                self.sheet_list = [sheet for sheet in self.df.values()]

            print(self.sheet_list[0].shape)
        except:
            log("cannot open xlsx!")
            traceback.print_exc()
            self._doc.error_code = [-3]

    def convert(self):
        self.init_package()
        if self._doc.error_code is not None:
            return

        sheet_no = 0
        for sheet in self.sheet_list:
            self._page = _Page(None, sheet_no)
            self.convert_page(sheet)

            if self._doc.error_code is None and self._page.error_code is not None:
                self._doc.error_code = self._page.error_code
            self._doc.add_child(self._page)
            sheet_no += 1

    def convert_page(self, sheet):
        text = '<table border="1">' + "\n"

        # 剔除多余空列
        max_row_len = 0
        max_col_len = 0
        if self.re_read:
            for index, row in sheet.iterrows():
                col_len = 0
                row_empty_flag = 1
                for i in range(len(row)):
                    if row[i] not in [None, "", np.nan]:
                        row_empty_flag = 0
                        col_len = i

                if self.re_read == 3 or self.re_read == 1:
                    if col_len > max_col_len:
                        max_col_len = col_len

                if self.re_read == 3 or self.re_read == 2:
                    if row_empty_flag == 0:
                        max_row_len = index

        for index, row in sheet.iterrows():
            if self.re_read == 3 or self.re_read == 2:
                if index > max_row_len:
                    break
            text = text + "<tr>"
            if self.re_read == 3 or self.re_read == 1:
                row = row[:max_col_len+1]
            for r in row:
                text = text + "<td>" + str(r) + "</td>" + "\n"
                # print(text)
            text = text + "</tr>" + "\n"
        text = text + "</table>" + "\n"

        _table = _Table(text, (0, 0, 0, 0), is_html=True)
        self._page.add_child(_table)

    def get_html(self):
        try:
            self.convert()
        except:
            traceback.print_exc()
            self._doc.error_code = [-1]
        if self._doc.error_code is not None:
            return self._doc.error_code
        return self._doc.get_html()