linux - xlsx 与 7z 的压缩方法

Question

我正在尝试以编程方式修改 excel 文件 (xlsx)。我可以成功解压，根据需要修改xml，然后重新压缩。但是，每次我打开 excel 时都会收到警告，即使它确实读取了文件。我相信错误是由于使用的压缩方法造成的。这是我能得到的最接近的示例:

解压

7z x original.xlsx -o./decomp_xlsx

..做一些事情..

压缩

7z a -tzip new ./decomp_xlsx/*

重命名

mv ./new.zip ./new.xlsx

我得到的错误是:Excel 在“new.xlsx”中发现不可读的内容。是否要恢复此工作簿的内容？如果您信任此工作簿的来源，请单击"is"。

来自 ECMA-376-2 Office 开放格式第 2 部分(打包约定)支持的压缩算法是 DEFLATE，如 .ZIP 规范中所述。包实现者不得使用除 DEFLATE 之外的任何压缩算法。

那么，我需要在 7z 或其他 linux 兼容程序中使用哪些开关才能在没有警告的情况下完成工作？我试过删除 -tzip 并使用 -m0=COPY，但 excel 甚至无法从那个中恢复。

这是 zip 程序和 zipinfo 的结果。我猜我不会找到一个工具来做这件事，除了下面提供的那个，所以我要奖励那个答案，看看我是否能找到人翻译成 python 进行测试。我不确定它是否处理了 4.5/3.0、b-/tx 或 defS/defF 之间的差异。

$ zipinfo original.xlsx
Archive:  original.xlsx
Zip file size: 228039 bytes, number of entries: 20
-rw----     4.5 fat     1969 b- defS 80-Jan-01 00:00 [Content_Types].xml
-rw----     4.5 fat      588 b- defS 80-Jan-01 00:00 _rels/.rels
-rw----     4.5 fat     1408 b- defS 80-Jan-01 00:00 xl/_rels/workbook.xml.rels
-rw----     4.5 fat      908 b- defS 80-Jan-01 00:00 xl/workbook.xml
-rw----     4.5 fat    35772 b- defS 80-Jan-01 00:00 xl/worksheets/sheet4.xml
-rw----     4.5 fat      322 b- defS 80-Jan-01 00:00 xl/worksheets/_rels/sheet4.xml.rels
-rw----     4.5 fat      322 b- defS 80-Jan-01 00:00 xl/worksheets/_rels/sheet1.xml.rels
-rw----     4.5 fat   230959 b- defS 80-Jan-01 00:00 xl/worksheets/sheet2.xml
-rw----     4.5 fat   263127 b- defS 80-Jan-01 00:00 xl/worksheets/sheet3.xml
-rw----     4.5 fat   295775 b- defS 80-Jan-01 00:00 xl/worksheets/sheet1.xml
-rw----     4.5 fat     1947 b- defS 80-Jan-01 00:00 xl/sharedStrings.xml
-rw----     4.5 fat    22698 b- defS 80-Jan-01 00:00 xl/styles.xml
-rw----     4.5 fat     7079 b- defS 80-Jan-01 00:00 xl/theme/theme1.xml
-rw----     4.5 fat      220 b- defS 80-Jan-01 00:00 xl/printerSettings/printerSettings2.bin
-rw----     4.5 fat   464247 b- defS 80-Jan-01 00:00 xl/externalLinks/externalLink1.xml
-rw----     4.5 fat      338 b- defS 80-Jan-01 00:00 xl/externalLinks/_rels/externalLink1.xml.rels
-rw----     4.5 fat      220 b- defS 80-Jan-01 00:00 xl/printerSettings/printerSettings1.bin
-rw----     4.5 fat      593 b- defS 80-Jan-01 00:00 docProps/core.xml
-rw----     4.5 fat    62899 b- defS 80-Jan-01 00:00 xl/calcChain.xml
-rw----     4.5 fat     1031 b- defS 80-Jan-01 00:00 docProps/app.xml
20 files, 1392422 bytes uncompressed, 223675 bytes compressed:  83.9%

$ zipinfo new.xlsx
Archive:  new.xlsx
Zip file size: 233180 bytes, number of entries: 20
-rw-r--r--  3.0 unx     1031 tx defF 80-Jan-01 00:00 docProps/app.xml
-rw-r--r--  3.0 unx      593 tx defF 80-Jan-01 00:00 docProps/core.xml
-rw-r--r--  3.0 unx    62899 tx defF 80-Jan-01 00:00 xl/calcChain.xml
-rw-r--r--  3.0 unx   464247 tx defF 80-Jan-01 00:00 xl/externalLinks/externalLink1.xml
-rw-r--r--  3.0 unx      338 tx defF 80-Jan-01 00:00 xl/externalLinks/_rels/externalLink1.xml.rels
-rw-r--r--  3.0 unx      220 bx defF 80-Jan-01 00:00 xl/printerSettings/printerSettings1.bin
-rw-r--r--  3.0 unx      220 bx defF 80-Jan-01 00:00 xl/printerSettings/printerSettings2.bin
-rw-r--r--  3.0 unx     1947 tx defF 80-Jan-01 00:00 xl/sharedStrings.xml
-rw-r--r--  3.0 unx    22698 tx defF 80-Jan-01 00:00 xl/styles.xml
-rw-r--r--  3.0 unx     7079 tx defF 80-Jan-01 00:00 xl/theme/theme1.xml
-rw-r--r--  3.0 unx      908 tx defF 80-Jan-01 00:00 xl/workbook.xml
-rw-r--r--  3.0 unx   295775 tx defF 80-Jan-01 00:00 xl/worksheets/sheet1.xml
-rw-r--r--  3.0 unx   230959 tx defF 80-Jan-01 00:00 xl/worksheets/sheet2.xml
-rw-r--r--  3.0 unx   263127 tx defF 80-Jan-01 00:00 xl/worksheets/sheet3.xml
-rw-r--r--  3.0 unx    35772 tx defF 80-Jan-01 00:00 xl/worksheets/sheet4.xml
-rw-r--r--  3.0 unx      322 tx defF 80-Jan-01 00:00 xl/worksheets/_rels/sheet1.xml.rels
-rw-r--r--  3.0 unx      322 tx defF 80-Jan-01 00:00 xl/worksheets/_rels/sheet4.xml.rels
-rw-r--r--  3.0 unx     1408 tx defF 80-Jan-01 00:00 xl/_rels/workbook.xml.rels
-rw-r--r--  3.0 unx     1969 tx defF 80-Jan-01 00:00 [Content_Types].xml
-rw-r--r--  3.0 unx      588 tx defF 80-Jan-01 00:00 _rels/.rels
20 files, 1392422 bytes uncompressed, 229608 bytes compressed:  83.5%

Answer 1

出于某种奇怪的原因，微软正在本地文件头和中央目录头中查看“需要提取的版本”中的操作系统编码。它希望这些为零，但 7z 将它们设置为 3 for Unix。如果您打算使用 7z，则需要修补生成的文件。

这个程序会这样做:

/* needz.c - zero the operating system byte for "version needed to extract" in
   the local and central headers of the zip files given on the command line.
   Placed in the public domain by Mark Adler, 23 Feb 2013. */

#include <stdio.h>
#include <stdlib.h>

static void bail(char *why, char *what)
{
    fprintf(stderr, "needz error: %s%s\n", why, what);
    exit(1);
}

/* Read len bytes from offset as a little-endian integer.  Negative offsets are
   considered to be from the end of the file. */
static unsigned long peek(FILE *stream, off_t offset, int len)
{
    int ret, shift;
    unsigned long val;

    ret = fseeko(stream, offset, offset < 0 ? SEEK_END : SEEK_SET);
    if (ret)
        bail("not a zip file", "");
    val = 0;
    shift = 0;
    while (len--) {
        ret = getc(stream);
        if (ret == EOF)
            bail("not a zip file", "");
        val += (unsigned long)ret << shift;
        shift += 8;
    }
    return val;
}

/* Write len bytes to offset from val as a little-endian integer.  Negative
   offsets are considered to be from the end of the file. */
static void poke(FILE *stream, off_t offset, int len, unsigned long val)
{
    int ret;

    ret = fseeko(stream, offset, offset < 0 ? SEEK_END : SEEK_SET);
    if (ret)
        bail("not a zip file", "");
    while (len--) {
        ret = putc(val, stream);
        if (ret == EOF)
            bail("could not write", "");
        val >>= 8;
    }
}

/* Zero out the OS byte in the extract fields.  This assumes the classic zip
   format (not Zip64), and no zip file comment. */
static void zip_zero_os(char *path)
{
    FILE *zip;
    unsigned entries;
    off_t central, local;

    zip = fopen(path, "r+b");
    if (zip == NULL)
        bail("could not open", path);
    if (peek(zip, -22, 4) != 0x06054b50)
        bail(path, " is not a zip file or has an end comment");
    entries = peek(zip, -12, 2);
    central = peek(zip, -6, 4);
    while (entries--) {
        if (peek(zip, central, 4) != 0x02014b50)
            bail(path, " has a structure error or is Zip64");
        poke(zip, central + 7, 1, 0);
        local = peek(zip, central + 42, 4);
        if (peek(zip, local, 4) != 0x04034b50)
            bail(path, " has a structure error or is Zip64");
        poke(zip, local + 5, 1, 0);
        central += 46 + peek(zip, central + 28, 2) +
                   peek(zip, central + 30, 2) + peek(zip, central + 32, 2);
    }
    if (fclose(zip) == EOF)
        bail("could not close ", path);
}

int main(int argc, char **argv)
{
    while (--argc)
        zip_zero_os(*++argv);
    return 0;
}