在 html 报告中可视化差异

Question

我想生成一份 html 报告，该报告可视化许多非常相似的文本之间的差异。我找到了 diffobj 包，如果在交互式会话中使用，它可以满足我的需求：

但是，我不知道如何在报告中呈现它。到目前为止，这是我尝试过的：

---
title: "Repex"
output: html_document
---

```{r cars}
duplicates <- data.frame(text_original = "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Maecenas laoreet risus et eros sagittis aliquam. Donec fringilla pharetra vestibulum. Fusce vestibulum imperdiet nibh ac rutrum. Aenean sollicitudin, tellus sed tempor varius, quam dolor ornare sapien, eu faucibus quam arcu vestibulum velit. Praesent maximus odio magna, in vulputate arcu cursus vitae. Praesent condimentum purus sit amet nisl vestibulum semper. Nunc quis eros ultricies, elementum eros sed, ullamcorper nunc. Nunc dictum commodo quam, et venenatis velit porta sit amet. Nunc et lorem et odio scelerisque vulputate sed at purus. Sed velit ipsum, consequat vel tristique tincidunt, semper in odio. Nullam pharetra laoreet velit quis sollicitudin. Fusce tellus felis, scelerisque id ipsum et, varius iaculis erat. Sed porttitor at quam sed rhoncus. Donec rutrum justo nec malesuada aliquam. Maecenas feugiat odio ac ante consequat, aliquet tempus magna tempus. Morbi convallis orci felis, ac ultricies ex dignissim in. Donec ornare vehicula ante eu interdum.",
                         text_duplicate = "Lorem dolor sit amet, consectetur elit. Maecenas laoreet risus et eros sagittis aliquam. Donec fringilla pharetra vestibulum. Fusce vestibulum imperdiet nibh ac rutrum. Aenean sollicitudin, tellus sed tempor varius, quam dolor ornare sapien, eu faucibus quam arcu vestibulum velit. Praesent maximus odio magna, in vulputate arcu cursus vitae. Praesent condimentum purus sit amet nisl vestibulum semper. Nunc quis eros ultricies, elementum sed, ullamcorper nunc. Nunc commodo quam, et venenatis velit porta sit amet. Nunc et lorem et odio scelerisque vulputate sed at purus. Sed velit ipsum, consequat vel tristique tincidunt, semper in odio. Nullam pharetra laoreet velit quis sollicitudin. Fusce tellus felis, scelerisque id ipsum et, varius iaculis erat. Sed porttitor at quam sed rhoncus. Donec rutrum justo nec malesuada aliquam. Maecenas feugiat odio ac ante consequat, aliquet tempus magna tempus. Morbi convallis orci felis, ac ultricies ex dignissim in. Donec ornare vehicula ante eu interdum.",
                         stringsAsFactors = FALSE)

library(diffobj)
for (i in 1) {
  orig <- unname(unlist(quanteda::tokens(duplicates$text_original[i], what = "sentence")))
  dup <- unname(unlist(quanteda::tokens(duplicates$text_duplicate[i], what = "sentence")))
  diff <- diffPrint(current = orig,
                    target = dup,
                    format = "html",
                    interactive = FALSE)
  print(diff)
}
```

编织此文件时的结果是显示每个差异的 html 代码但未呈现。请注意，我仅使用一次迭代的 for 循环来进行演示。真正的任务是渲染大约 50 个差异。

有什么办法可以通过调整 diffPrint() 或 rmarkdown 的设置来解决这个问题吗？

Answer 1

问题是 diffPrint 不输出 css 样式。

在设置块中，添加结果"asis"

```{r cars, results="asis"}
duplicates <- data.frame(text_original = "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Maecenas laoreet risus et eros sagittis aliquam. Donec fringilla pharetra vestibulum. Fusce vestibulum imperdiet nibh ac rutrum. Aenean sollicitudin, tellus sed tempor varius, quam dolor ornare sapien, eu faucibus quam arcu vestibulum velit. Praesent maximus odio magna, in vulputate arcu cursus vitae. Praesent condimentum purus sit amet nisl vestibulum semper. Nunc quis eros ultricies, elementum eros sed, ullamcorper nunc. Nunc dictum commodo quam, et venenatis velit porta sit amet. Nunc et lorem et odio scelerisque vulputate sed at purus. Sed velit ipsum, consequat vel tristique tincidunt, semper in odio. Nullam pharetra laoreet velit quis sollicitudin. Fusce tellus felis, scelerisque id ipsum et, varius iaculis erat. Sed porttitor at quam sed rhoncus. Donec rutrum justo nec malesuada aliquam. Maecenas feugiat odio ac ante consequat, aliquet tempus magna tempus. Morbi convallis orci felis, ac ultricies ex dignissim in. Donec ornare vehicula ante eu interdum.",
                         text_duplicate = "Lorem dolor sit amet, consectetur elit. Maecenas laoreet risus et eros sagittis aliquam. Donec fringilla pharetra vestibulum. Fusce vestibulum imperdiet nibh ac rutrum. Aenean sollicitudin, tellus sed tempor varius, quam dolor ornare sapien, eu faucibus quam arcu vestibulum velit. Praesent maximus odio magna, in vulputate arcu cursus vitae. Praesent condimentum purus sit amet nisl vestibulum semper. Nunc quis eros ultricies, elementum sed, ullamcorper nunc. Nunc commodo quam, et venenatis velit porta sit amet. Nunc et lorem et odio scelerisque vulputate sed at purus. Sed velit ipsum, consequat vel tristique tincidunt, semper in odio. Nullam pharetra laoreet velit quis sollicitudin. Fusce tellus felis, scelerisque id ipsum et, varius iaculis erat. Sed porttitor at quam sed rhoncus. Donec rutrum justo nec malesuada aliquam. Maecenas feugiat odio ac ante consequat, aliquet tempus magna tempus. Morbi convallis orci felis, ac ultricies ex dignissim in. Donec ornare vehicula ante eu interdum.",
                         stringsAsFactors = FALSE)

library(diffobj)
for (i in 1) {
  orig <- unname(unlist(quanteda::tokens(duplicates$text_original[i], what = "sentence")))
  dup <- unname(unlist(quanteda::tokens(duplicates$text_duplicate[i], what = "sentence")))
  diff <- diffPrint(current = orig,
                    target = dup,
                    format = "html",
                    interactive = FALSE)
  print(diff)
}
```

在rmd文件的某处添加

<style type="text/css">
DIV.diffobj_container PRE {
  white-space: pre-wrap;
  margin: 0;
}
DIV.diffobj_container DIV.row {
  width: 100%;
  font-family: monospace;
  display: table;
  table-layout: fixed;
}
DIV.diffobj_container DIV.line {
  width: auto;
  display: table-cell;
  overflow: hidden;
}
DIV.diffobj_container DIV.line>DIV {
  width: 100%;
  display: table;
  table-layout: auto;
}
DIV.diffobj_container DIV.line.banner>DIV {
  display: table;
  table-layout: auto; /* set to fixed in JS */
}
DIV.diffobj_container DIV.text {
  display: table-cell;
  width: 100%;
}
DIV.diffobj_container DIV.gutter {
  display: table-cell;
  padding: 0 0.2em;
}
DIV.diffobj_container DIV.gutter DIV {
  display: table-cell;
}
#diffobj_content_meta DIV.diffobj_container DIV.row {
  width: auto;
}
#diffobj_banner_meta DIV.diffobj_container DIV.line.banner>DIV {
  table-layout: auto;
}
#diffobj_outer {
  overflow: hidden;
}
/* Summary -------------------------------------------------------------------*/ 

DIV.diffobj_container DIV.summary DIV.map {
  word-wrap: break-word;
  padding-left: 1em;
}
DIV.diffobj_container DIV.summary DIV.detail {
  padding-left: 1em;
}

/* Common elements -----------------------------------------------------------*/

DIV.diffobj_container DIV.line.banner {
  font-size: 1.2em;
  font-weight: bold;
  overflow: hidden;
}
/* truncate banners */
DIV.diffobj_container DIV.line.banner DIV.text DIV{
  white-space: nowrap;
  overflow: hidden;
  text-overflow: ellipsis;
  width: 100%;             /* need to compute and set in JS */
}
DIV.diffobj_container DIV.gutter,
DIV.diffobj_container DIV.guide,
DIV.diffobj_container DIV.fill,
DIV.diffobj_container DIV.context_sep,
DIV.diffobj_container SPAN.trim {
  color: #999;
}
DIV.diffobj_container DIV.header {
  font-size: 1.1em;
}
DIV.diffobj_container DIV.text>DIV.match,
DIV.diffobj_container DIV.text>DIV.guide {
  background-color: #ffffff;
}
DIV.diffobj_container DIV.text>DIV.fill {
  background-color: transparent;
}
DIV.diffobj_container DIV.text>DIV {
  padding-right: 3px;
}
DIV.diffobj_container DIV.text>DIV {
  border-left: 1px solid #888888;
}
DIV.diffobj_container DIV.line {
  background-color: #eeeeee;
}
DIV.diffobj_container DIV.text>DIV,
DIV.diffobj_container DIV.header {
  padding-left: 0.5em;
}
DIV.diffobj_container DIV.line>DIV.match,
DIV.diffobj_container DIV.line>DIV.fill,
DIV.diffobj_container DIV.line>DIV.guide {
  border-left: 1px solid #888888;
}
/* github inspired color scheme - default ------------------------------------*/

DIV.diffobj_container.light.rgb SPAN.word.insert,
DIV.diffobj_container.light.rgb DIV.line>DIV.insert {
  background-color: #a6f3a6;
}
DIV.diffobj_container.light.rgb SPAN.word.delete,
DIV.diffobj_container.light.rgb DIV.line>DIV.delete {
  background-color: #f8c2c2;
}
DIV.diffobj_container.light.rgb DIV.text>DIV.insert {
  background-color: #efffef;
}
DIV.diffobj_container.light.rgb DIV.text>DIV.insert,
DIV.diffobj_container.light.rgb DIV.line>DIV.insert {
  border-left: 1px solid #33bb33;
}
DIV.diffobj_container.light.rgb DIV.text>DIV.delete {
  background-color: #ffefef;
}
DIV.diffobj_container.light.rgb DIV.text>DIV.delete,
DIV.diffobj_container.light.rgb DIV.line>DIV.delete {
  border-left: 1px solid #cc6666;
}
DIV.diffobj_container.light.rgb DIV.header {
  background-color: #e0e6fa;
  border-left: 1px solid #9894b6;
}
/* Yellow Blue variation -----------------------------------------------------*/

DIV.diffobj_container.light.yb SPAN.word.insert,
DIV.diffobj_container.light.yb DIV.line>DIV.insert {
  background-color: #c0cfff;
}
DIV.diffobj_container.light.yb SPAN.word.delete,
DIV.diffobj_container.light.yb DIV.line>DIV.delete {
  background-color: #e7e780;
}
DIV.diffobj_container.light.yb DIV.text>DIV.insert {
  background-color: #efefff;
}
DIV.diffobj_container.light.yb DIV.text>DIV.insert,
DIV.diffobj_container.light.yb DIV.line>DIV.insert {
  border-left: 1px solid #3333bb;
}
DIV.diffobj_container.light.yb DIV.text>DIV.delete {
  background-color: #fefee5;
}
DIV.diffobj_container.light.yb DIV.text>DIV.delete,
DIV.diffobj_container.light.yb DIV.line>DIV.delete {
  border-left: 1px solid #aaaa55;
}
DIV.diffobj_container.light.yb DIV.header {
  background-color: #afafaf;
  border-left: 1px solid #e3e3e3;
  color: #e9e9e9;
}
DIV.diffobj_container.light.yb DIV.line {
  background-color: #eeeeee;
}
</style>

A working example

source rmarkdown

郑重声明，related discussion on github

在 html 报告中可视化差异

Visualise diff in html report

r

r-markdown

diffobj