[Blog] Part4: Blog with RStudio, R, RMarkdown, Jekyll and Github
一直以来都想构建一个比较完整又自动化的个人博客站。这里有两个关键字，即完整和自动。怎么说呢，如果我只是一个单纯的文字工作者，写博文用简书足够，简洁的排版，你只需负责创作。但是如果又需要代码，可以考虑CSDN，专注技术博文。看回我的需求，希望代码和文字一体，且代码运行结果能自动显示，写完.Rmd自动转换成.md，图片自动变为Markdown可接受的Link，之后同步上传到github进行安全托管。
注: .md是Github Pages上唯一接受的file extension.
这篇博文是[GitHub 博客简明使用指南]与[技术札记]的衍生，或许可以综合成一个系列，让小白少走些弯路。
用过Rstudio的RUsers都知道，左上角的R Notebook已经可以完成代码与结果一体化的需求，但是要将运行完成的.Rmd文件转化为.md格式，并自动显示图片或结果，就需要用到额外的函数。
以下是我根据其他作者编写进行修改的rmd2md.R函数，运行函数前在博客官网目录下创建三个文件夹，包括：_posts, _rmd, pictures。
写函数前，确定必要的文件路径
rmd2md <- function( path_site = getwd(), # 当前博客地址
dir_rmd = "_rmd", # _rmd文件夹地址
dir_md = "_posts", # _posts文件夹地址
#dir_images = "figures",
url_images = "figures/",
out_ext='.md',
in_ext='.Rmd',
recursive=FALSE)
加载包
# Loading the necessary packages
require(knitr, quietly=TRUE, warn.conflicts=FALSE)
找到需要转化的.Rmd文件
#andy change to avoid path problems when running without sh on windows
# list out each rmd path in _rmd file using 'list.files'
files <- list.files(path=file.path(path_site,dir_rmd), pattern=in_ext, ignore.case=TRUE, recursive=recursive)
# for each file in files
for(f in files) {
# pop up the processing message
message(paste("Processing ", f, sep=''))
对每个文件进行文本搜寻
- 判断此文件是否已转换: 定位‘—‘，确定两个的位置，第一条在首行，第二条在两行之后，若找不到弹出warning
- 判断Status的位置：确定在‘—’中间后取得状态内容，调整大小写，判断是否处理过，若处理过弹出warning；若没有则转化成处理了，并进行knit转化（使用try-else; Python: try-except）
# determine the file and publish status
content <- readLines(file.path(path_site,dir_rmd,f))
frontMatter <- which(substr(content, 1, 3) == '---')
if(length(frontMatter) >= 2 & 1 %in% frontMatter) {
statusLine <- which(substr(content, 1, 7) == 'status:')
publishedLine <- which(substr(content, 1, 10) == 'published:')
if(statusLine > frontMatter[1] & statusLine < frontMatter[2]) {
status <- unlist(strsplit(content[statusLine], ':'))[2]
status <- sub('[[:space:]]+$', '', status)
status <- sub('^[[:space:]]+', '', status)
if(tolower(status) == 'process') {
#This is a bit of a hack but if a line has zero length (i.e. a
#black line), it will be removed in the resulting markdown file.
#This will ensure that all line returns are retained.
content[nchar(content) == 0] <- ' '
message(paste('Processing ', f, sep=''))
content[statusLine] <- 'status: publish'
content[publishedLine] <- 'published: true'
#andy change to path
outFile <- file.path(path_site, dir_md, paste0(substr(f, 1, (nchar(f)-(nchar(in_ext)))), out_ext))
#render_markdown(strict=TRUE)
#render_markdown(strict=FALSE) #code didn't render properly on blog
#andy change to render for jekyll
render_jekyll(highlight = "pygments")
#render_jekyll(highlight = "prettify") #for javascript
opts_knit$set(out.format='markdown')
# andy BEWARE don't set base.dir!! it caused me problems
# "base.dir is never used when composing the URL of the figures; it is
# only used to save the figures to a different directory.
# The URL of an image is always base.url + fig.path"
# https://groups.google.com/forum/#!topic/knitr/18aXpOmsumQ
opts_knit$set(base.url = "/")
opts_chunk$set(fig.path = url_images)
#andy I could try to make figures bigger
#but that might make not work so well on mobile
#opts_chunk$set(fig.width = 8.5,
# fig.height = 5.25)
try(knit(text=content, output=outFile), silent=FALSE)
} else {
warning(paste("Not processing ", f, ", status is '", status,
"'. Set status to 'process' to convert.", sep=''))
}
} else {
warning("Status not found in front matter.")
}
} else {
warning("No front matter found. Will not process this file.")
}
}
invisible()
}