Introduction
Sometimes when I write technical documents or notes by myself, I find that the inserted pictures under the md syntax are either local or image beds.
The disadvantage of local is slow loading on the server (gitgub), or forgetting to post pictures when sending documents to others. . .
I
have seen pictures directly embedded in the document before. Of course, the disadvantage is that it will increase the file size, but if the picture is lost,
some people will say, a The bs64 code converted from a few megabytes of pictures is longer than my article (I don’t believe you, try copying the bs64 below to see how many characters, Xiaoxin crashes when copying -.-), is there any way? ?
There must be a way
to number the pictures, and put all the pictures base64 encoding at the end of the document. The format is as follows:
where the picture is inserted: ！[图片缺失描述][图片编号]
at the end of the document: [图片编号]:data:image/jpg;base64,base64编码
but there is another disadvantage that Typora will get stuck. Said it could not be opened. . Actually less than 4M
! ! ! This is a headache for me, so I thought about it for a while. The image that was not compressed before and stored in the md file is still relatively large, so I studied py compression.
effect
format! ! !
Where the picture is inserted: ！[图片缺失描述][图片编号]
at the end of the document:[图片编号]:data:image/jpg;base64,base64编码
Online conversion
Bing search pictures directly to bs64,
such as here
py convert bs64
import base64
f=open('lady.jpg','rb') #图文件lady.jpg，自行更改
pbs64=base64.b64encode(f.read()) #读取文件内容，转换为base64编码
f.close()
print(pbs64)
In this way, expanding the bs64 code py will be stuck (2k lines...), so I make a slight improvement, and write the generated large byte stream into a text file (bs64.txt), and open it with Notepad. Dead, the notepad stuck for two seconds in the actual test, editplus is much better, there is basically no such problem, and bs64 in editplus does not automatically wrap by default, it is very comfortable to copy a line directly
import base64
f=open('lady.jpg','rb') #图文件lady.jpg，自行更改
pbs64=base64.b64encode(f.read()) #读取文件内容，转换为base64编
f.close()
#print(pbs64)
mf=open("bs64.txt","wb") #新建目标文件存储bs64码
mf.write(pbs64)
mf.close
You ask me why I don’t put an example here? ? ? Because CSDN can't bear to copy the bs64 code. . . . If you don’t believe
me , look at the end of the article, [I originally copied it here to embed bs64, but it stuck. . Can not save or publish articles]
example look at my blog site it
http://kearney.club/2020/06/11/%E6%88%91%E7%9A%84%E8%97%8F%E5%B0% B8%E4%B8%8D%E8%A7%81%E4%BA%86/
This is static hosting on github Page, which is displayed normally. Actually, my article is more than a thousand words in the editing process of Typora, and three pictures are added The bs64 becomes 3w multi-word. . . I looked at the background and the memory occupied is not large, but it may be the problem of the amount of data loaded by typora itself. Copying and pasting in the 3w multi-character is a bit lagging (I have more than 60% of the free 32G memory).
lossless compression
How to use: Pull directly to the bottom of the code and modify the picturepath. The
picture and code need to be in the same directory
省略的代码放在了github上。。省的大家看得烦
https://github.com/BackMountainDevil/image-into-bs64
def img_bs64(infile,outfile='', savefile =''):
compress_image(infile)
outfile = get_outfile(infile, outfile)
f=open(outfile,'rb')
pbs64=base64.b64encode(f.read())
f.close()
dir, suffix = os.path.splitext(infile)
savefile = '{}-out{}'.format(dir, '.txt')
mf=open(savefile,"wb")
mf.write(pbs64)
mf.close
os.remove(outfile)
if __name__ == '__main__':
picturepath= "luffy.jpg"
img_bs64(picturepath)
The code is all cv. Do you really leave without a thumbs up?? It’s OK to click a Star
at Github
reference
- https://github.com/BackMountainDevil/image-into-bs64
- https://cn.bing.com/search?q=%e5%9b%be%e7%89%87%e8%bd%acbs64&qs=SC&pq=%e5%9b%be%e7%89%87%e8%bd%acbs&sc=1-5&cvid=5B916F1601B6423BBAC2240FC489E031&FORM=QBRE&sp=1
- https://oktools.net/image2base64
- https://blog.csdn.net/baidu_31492511/article/details/90704123
- https://blog.csdn.net/cxs123678/article/details/82533306
- https://www.jianshu.com/p/280c6a6f2594
- https://www.runoob.com/python/python-files-io.html
- https://blog.csdn.net/baidu_33221362/article/details/81566874
- https://blog.csdn.net/qq_31567335/article/details/82322858?utm_medium=distribute.pc_relevant.none-task-blog-BlogCommendFromMachineLearnPai2-2.add_param_isCf&depth_1-utm_source=distribute.pc_relevant.none-task-blog-BlogCommendFromMachineLearnPai2-2.add_param_isCf
- https://www.cnblogs.com/li1992/p/10675769.html
- https://blog.csdn.net/julac/article/details/105492336
- https://github.com/BackMountainDevil/image-into-bs64
- https://blog.csdn.net/qq_36387683/article/details/100579736
![picture][1]
其它正文
末尾
[1]:data:image/jpg;base64,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