csdn文章分类修改脚本

以前写的文章分类太多，太乱了，所以决定来一次清理，把一些文章类别给替换掉。和LX同学讨论了一个下午，终于有一个方案了，搞了一晚上才弄好。

我是混合了python和js代码一起做的。js可以在浏览器的 控制窗下运行，不用登陆了，所以很方便。python我用的比较习惯，所以处理数据比较方便，而且不会丢失。

 

步骤如下：

step 1：先把文章列表抓下来，知道有哪些文章。markdown编写的文章不能处理！都进eid了！

 

1. import urllib.request

2.  

3. import re

4. ids = {} #保存所有的文章id

5.  

6. for i in range(30):#按15篇文章一页算自己有多少页

7. url = "http://blog.csdn.net/firenet1/article/list/"+str(i+1)

8. try:

9. data = urllib.request.urlopen(url).read()

10. except :

11. continue

12. data = data.decode('UTF-8')

13. patterm = "/firenet1/article/details/[0-9]{5,10}"

14. patterm = re.compile(patterm)

15. data = patterm.findall(data,re.S|re.M|re.I)

16. digit = re.compile(r"[0-9]{5,10}",re.S|re.M|re.I)

17. print(len(data))

18. for d in data:

19. # print(d)

20. d = digit.search(d)

21. # print(d)

22. # print(d.group())

23. ids[d.group()] = 1

24. print(len(ids))

25. file = open("js_array.txt","w")

26. out = "var id_array = new Array("

27. for i in ids.keys():

28. out+="\""+i+"\",\n"

29. out+=");"

30. file.write(out)

31. file.close()

32. '''

33. 输出文件保存成js数组

34. var id_array = new Array("77187506",

35. "77073144",

36. "77046721",

37. "76766916",

38. "76642319",

39. "76195994")

40. '''

step 2： 打开一个csdn博客编辑页，任何一页都行。F12进入控制台，在控制台里执行以下代码，这一步就能获得大部分文章的id，tag，类别

 

 

1. var id_array = new Array("77187506", //这个就是上一步得到的文章id数组

2. "77073144",

3. "77046721",

4. );

5. #抓取文章标签和类别

6. var time_out_th = 500;

7. var i;

8. var a;

9. var id;

10. var out;

11. var eid; //执行不成功的文章id记录下来，可以自己打印出来看

12. function getdata(){ //调用函数入口，通过timeout设置，每次抓取一个文章才进行下一个文章抓取

13. i = 0;

14. out = new Array();

15. id = id_array[i];

16. eid = new Array();

17. a = window.open(id,"_blank"); //打开新的编辑页面

18. setTimeout(doing,time_out_th);

19. }

20. function doing(){

21. if(i == id_array.length) {

22. console.log("finish");

23. return i;

24. }

25. try{

26. if(a != 0 && a.document.readyState == "complete"){ //判断新打开的页面是不是加载完毕了

27. let cla = a.document.getElementById("txtTag").value;

28. let lable = a.document.getElementById("d_tag2").innerHTML;

29. lable = lable.replace("\n","");

30. out.push(new Array(id,cla,lable));

31. a.close();

32. console.log("ok: "+i)

33. i++;

34. if(i == id_array.length) {

35. console.log("finish");

36. return i;

37. }

38. id = id_array[i];

39. a = window.open(id,"_blank");

40. }

41. }

42. catch(err){

43. eid.push(id);

44. console.log("err: "+i);

45. i++;

46. a.close();

47. if(i == id_array.length) {

48. console.log("finish");

49. return i;

50. }

51. id = id_array[i];

52. a = window.open(id,"_blank");

53. }

54. setTimeout(doing,time_out_th);

55. }

56. function output(){ //输出文本函数

57. let res = "";

58. for(let i = 0;i < out.length;i++){

59. res += out[i][0]+"###,###"+out[i][1]+"###,###"+out[i][2]+"\n";

60. }

61. console.log(res);

62. }

63. /*

64. 输出数据如下：复制黏贴到本地文本后用于后面的步骤

65. 47311009###,###多校联合训练赛###,###<span title="单击删除该标签">hdu 5328</span><span title="单击删除该标签">hdu</span>

66. */

step 3：打开自己的类别管理，F12进入控制台，执行以下代码

 

 

1. ##获取文章类别总数

2. var x = document.getElementsByClassName("tdleft")

3. var y = ""

4. for(var i = 1;i < x.length; i++){

5. y += (x[i].firstChild.innerHTML)+"###,###\n";

6. }

7. console.log(y)

8. /**

9. 输出如下：每一行就是一个类别 ###,###以及后面的部分就是我想把这个类别换成其他类别,分割开,可以没有

10. 动态规划###,###ACM-ICPC编程题,动态规划

11. 数据结构###,###ACM-ICPC编程题,数据结构

12. 字符串###,###ACM-ICPC编程题,字符串

13. 模拟###,###ACM-ICPC编程题,模拟

14. */

step 4：把所有文章的类别和tag都换成新的，我的tag是使用原来的tag，如果不够5个，会把原先的类别变成tag。python代码

1. import urllib.request

2.  

3. import re

4. patterm = "<span title=\"单击删除该标签\">[\w]+</span>"

5. patterm = re.compile(patterm)

6. articles = []

7. #读取原来文章的tag和分类，并且处理成数组

8. with open("classify_lable.txt","r",encoding="utf-8") as f:

9. for i in f:

10. i = i.split("###,###")

11. i[2] = patterm.findall(i[2])

12. k = ""

13. j = 0

14. while j < len(i[2]):

15. i[2][j] = i[2][j].replace("<span title=\"单击删除该标签\">","")

16. i[2][j] = i[2][j].replace("</span>","")

17. k+=" "+i[2][j]

18. j+=1

19.  

20. articles.append(i)

21. print("articles: "+str(len(articles)))

22.  

23. #读取原来的分类列表，并且映射成字典，方便下一步

24. origin = {}

25. now = {}

26. with open("classfyMap.txt","r",encoding="utf-8") as f:

27. for i in f:

28. i = i.replace("\n","")

29. i = i.split("###,###")

30. if(len(i[1]) == 0):

31. continue

32. i[1] = i[1].split(",")

33. origin[i[0]] = i[1]

34. for j in i[1]:

35. now[j] = 1

36. print("old category: %d"%(len(origin)))

37. print("now category: %d"%(len(now)))

38. # for i in now.keys():

39. # print(i)

40. #对于每一篇文章，更新tag以及把旧的类别换成新的，新的类别以'##'开头

41. #这里用字典可以去重，tag和类别都不会重复

42. new_article = []

43. nolable = 0

44. nocategory = 0

45. for i in articles:

46. lable_c = {}

47. new_c = {}

48. for j in i[2]:

49. if(len(j) > 0):

50. lable_c[j] = 1

51. for j in i[1].split(","):

52. if(len(j) > 0):

53. lable_c[j] = 1

54. if j in origin:

55. for k in origin[j]:

56. new_c[k] = 1

57. category = []

58. lable = []

59. for j in lable_c.keys():

60. lable.append(j)

61. if(len(new_c) == 0):

62. for j in lable:

63. if(j in origin):

64. new_c[j] = 1

65. for j in new_c.keys():

66. category.append(j)

67. if(len(lable) == 0):

68. nolable += 1

69. if(len(category) == 0):

70. nocategory += 1

71. new_article.append([i[0],lable,category])

72. print("finale set: %d nolable: %d nocategory: %d"%(len(new_article),nolable,nocategory))

73. # for i in new_article:

74. # print(i)

75. #输出成js数组，用于下一步操作

76. js_arry = "var arti = ["

77. m = 0

78. for i in new_article:

79. lab = ""

80. cat = ""

81. for j in i[1]:

82. lab += ("<span title=\'单击删除该标签\'>%s</span>"%(j))

83. k = 0

84. while k < len(i[2]):

85. if k > 0:

86. cat +=','

87. cat += "##"+i[2][k]

88. k += 1

89. if (m > 0):

90. js_arry += ",\n"

91. js_arry += '["%s","%s","%s"]'%(i[0],lab,cat)

92.  
93.  

94. m+=1

95. js_arry +="];\n"

96. print(js_arry)

step 5：这一步就把js代码放到控制台运行了，控制台还是要编辑页面的，这样就没有域的问题。因为url我没处理哦！执行比较久300+文章呢

1. #更新文章类别和标签

2. var arti = [["77187506","<span title='单击删除该标签'>布隆过滤器</span><span title='单击删除该标签'>我只想找工作</span>","##我只想找工作"],

3. ["77073144","<span title='单击删除该标签'>hyperloglog</span><span title='单击删除该标签'>基数计数</span><span title='单击删除该标签'>我只想找工作</span>","##我只想找工作"],

4. ["77046721","<span title='单击删除该标签'>信号量</span><span title='单击删除该标签'>临界区</span><span title='单击删除该标签'>自旋锁</span><span title='单击删除该标签'>操作系统</span><span title='单击删除该标签'>我只想找工作</span>","##操作系统,##我只想找工作"],

5. 47438411","<span title='单击删除该标签'>2015多校联合训练赛</span><span title='单击删除该标签'>模拟</span>","##ACM-ICPC编程题,##模拟"],

6. ];

7. var time_out_th = 500;

8. var i;

9. var a;

10. var id;

11. var step = 0;

12. var eid = new Array();

13. function getdata(){

14. i = 0;

15. id = arti[i];

16. step = 0;

17. eid = new Array();

18. a = window.open(id[0],"_blank");

19. setTimeout(doing,time_out_th);

20. }

21. function doing(){

22. if(i == arti.length) {

23. console.log("finish");

24. return i;

25. }

26. try{//加载完成后修改内容并点击保存,延时200毫秒再判断是否保存好了

27. if(step == 0 && a != 0 && a.document.readyState == "complete"){

28. a.document.getElementById("txtTag").value = id[2];

29. a.document.getElementById("d_tag2").innerHTML = id[1];

30. console.log("complete: "+i)

31. step += 1

32. a.document.getElementById("btnDraft").click()

33. }

34. }

35. catch(err){

36. console.log(err);

37. console.log("err: "+i);

38. eid.push(id[0]);

39. i++;

40. a.close();

41. if(i == arti.length) {

42. console.log("finish");

43. return i;

44. }

45. step = 0

46. id = arti[i];

47. a = window.open(id[0],"_blank");

48. }

49. if(step == 1 && !a.saving){ //保存完毕，打开下一个网页

50. a.close();

51. console.log("ok: "+i)

52. i++;

53. if(i == arti.length) {

54. console.log("finish");

55. return i;

56. }

57. id = arti[i];

58. a = window.open(id[0],"_blank");

59. step = 0;

60. }

61. setTimeout(doing,time_out_th);

62. }

step 6：打开文章分类，把只有0篇的分类删除掉。alert手动确认，不然好像因为缓存的原因，删不掉后面的

1. #删除文章数为0的类别

2. var a = document.getElementsByClassName("red");

3. for(var i = 0;i < a.length;i++){

4. var b = a[i].href;

5. b = b.substring(b.length-7,b.length);

6. var c = a[i].text;

7. if(c == "0"){

8. console.log(b+" "+c);

9. $.get("?t=" + "del", { id: b, r: csdn.random() }, function (ret) {

10. alert(1);

11. });

12. //ory?t=del&id=1380215&r=83505

13. }

14. }

源代码参考：https://gitlab.com/linyuwang/csdn-classfy-change/tree/master
 

step 7：按文章数对类型进行排序。在文章类别管理的console执行

 

1. #把标签按文章数排序 这段代码有问题 总是有些没有移动的难道是缓存的问题？

2. #反复执行几次才行

3. var a = document.getElementsByClassName("red");

4. var ids = new Array();

5. var compare = function(x,y){

6. if(x[1] > y[1]) return -1;

7. if(x[1] == y[1]) return 0;

8. return 1;

9. }

10. for(var i = 0;i < a.length;i++){

11. var b = a[i].href;

12. b = b.substring(b.length-7,b.length);

13. var c = a[i].text;

14. ids.push(new Array(b,Number(c),i));

15. // console.log(b+" "+c);

16. // $.get("?t=" + "del", { id: b, r: csdn.random() }, function (ret) {

17. }

18. ids.sort(compare);

19. console.log(ids);

20.  

21. var i = 0

22. function doing(){

23. while(i != ids.length && ids[i][2] <= i){

24. i++;

25. }

26. if(i == ids.length) return "finish";

27. if(ids[i][2] > i){

28. ids[i][2]--;

29. let t = i;

30. // $.get("?t=" + "up", { id: ids[i][0], r: csdn.random() },function () {

31. // alert(t+","+ids[t][2]+","+ids[t][1]);});

32. doExec("", ids[i][0], "up");

33. }

34. console.log(i+","+ids[i][2]+","+ids[i][1]);

35. setTimeout(doing,500);

36. }

37. doing();