怎么使用Python中的正则表达式处理html文件
使用Python中的正则表达式处理html文件
finditer方法是一种全匹配方法。您可能已经使用了findall方法，它返回多个匹配字符串的列表。finditer返回一个迭代器顺序地为多个匹配中的每一个生成匹配对象。在下面的代码中，这些匹配对象被访问（通过for循环），因此可以打印组1。
您的任务是编写Python RE来识别HTML文本文件中的某些模式。将代码添加到STARTER脚本为这些模式编译RE（将它们分配给有意义的变量名称），并将这些RE应用于文件的每一行，打印出找到的匹配项。
1.编写识别HTML标签的模式，然后将其打印为“TAG:TAG string”（例如“TAG:b”代表标签）。为了简单起见，假设左括号和右括号每个标记的（<，>）将始终出现在同一行文本中。第一次尝试可能使regex“<.*>”其中“.”是与任何字符匹配的预定义字符类符号。尝试找出这一点，找出为什么这不是一个好的解决方案。编写一个更好的解决方案，解决这个问题
2.修改代码，使其区分开头和结尾标记（例如p与/p)打印OPENTAG和CLOSETAG
import sys, re
#------------------------------
testRE = re.compile('(logic|sicstus)', re.I)
testI = re.compile('<[A-Za-z]>', re.I)
testO = re.compile('<[^/](S*?)[^>]*>')
testC = re.compile(']*>')
with open('RGX_DATA.html') as infs:
linenum = 0
for line in infs:
linenum += 1
if line.strip() == '':
continue
print(' ', '-' * 100, '[%d]' % linenum, 'n TEXT:', line, end='')
m = testRE.search(line)
if m:
print('** TEST-RE:', m.group(1))
mm = testRE.finditer(line)
for m in mm:
print('** TEST-RE:', m.group(1))
index= testI.finditer(line)
for i in index:
print('Tag:',i.group().replace('<', '').replace('>', ''))
open1= testO.finditer(line)
for m in open1:
print('opening:',m.group().replace('<', '').replace('>', ''))
close1= testC.finditer(line)
for n in close1:
print('closing:',n.group().replace('<', '').replace('>', ''))请注意，有些HTML标签有参数，例如：
确保打开标记的模式适用于带参数和不带参数的标记，即成功找到并打印标签标签。现在扩展您的代码，以便打印两个打开的标签标签和参数，例如:
OPENTAG: table
PARAM: border=1
PARAM: cellspacing=0
PARAM: cellpadding=8
open1= testO.finditer(line)
for m in open1:
#print('opening:',m.group().replace('<', '').replace('>', ''))
firstm= m.group().replace('<', '').replace('>', '').split()
num = 0
for otherm in firstm:
if num == 0:
print('opening:',otherm)
else:
print('pram:',otherm)
num+= 1在正则表达式中，可以使用反向引用来指示匹配早期部分的子字符串,应再次出现正则表达式的。格式为N（其中N为正整数），并返回到第N个匹配的文本正则表达式组。例如，正则表达式，如：r" (w+) 1 仅当与组（w+）完全匹配的字符串再次出现时才匹配 backref1出现的位置。这可能与字符串“踢”匹配.例如，“the”出现两次。使用反向引用编写一个模式，当一行包含成对的open和关闭标签，例如在粗体中.
考虑到我们可能想要创建一个执行HTML剥离的脚本，即一个HTML文件，并返回一个纯文本文件，所有HTML标记都已从中删除出来这里我们不打算这样做，而是考虑一个更简单的例子，即删除我们在输入数据文件的任何行中找到的HTML标记。
你应该能够让您已经定义的RE识别HTML标签这样做,将生成的文本打印到屏幕上为STRIPPED：。。
import sys, re #------------------------------ # PART 1: # Key thing is to avoid matching strings that include # multiple tags, e.g. treating '' as a single # tag. Can do this in several ways. Firstly, use # non-greedy matching, so get shortest possible match # including the two angle brackets: tag = re.compile('') # The above treats the '/' of a close tag as a separate # optional component - so that this doesn't turn up as # part of the match '.group(1)', which is meant to return # the tag label. # Following alternative solution uses a negated character # class to explicitly prevent this including '>': tag = re.compile(']+)>') # Finally, following version separates finding the tag # label string from any (optional) parameters that might # also appear before the close angle bracket: tag = re.compile(r']+)?>') # Note that use of 'b' (as word boundary anchor) here means # we must mark the regex string as a 'raw' string (r'..'). #------------------------------ # PART 2: # Following closeTag definition requires first first char # after the open angle bracket to be '/', while openTag # definition excludes this by requiring first char to be # a 'word char' (w): openTag = re.compile(r'<(w[^>]*)>') closeTag = re.compile(r']*)>') # Following revised definitions are more carefully stated # for correct extraction of tag label (separately from # any parameters: openTag = re.compile(r'<(w+b)([^>]+)?>') closeTag = re.compile(r'') #------------------------------ # PART 3: # Above openTag definition will already get the string # encompassing any parameters, and return it as # m.group(2), i.e. defn: openTag = re.compile(r'<(w+b)([^>]+)?>') # If assume that parameters are continuous non-whitespace # chars separated by whitespace chars, then we can divide # them up using split - and that's how we handle them # here. (In reality, parameter strings can be a lot more # messy than this, but we won't try to deal with that.) #------------------------------ # PART 4: openCloseTagPair = re.compile(r'<(w+b)([^>]+)?>(.*?)') # Note use of non-greedy matching for the text falling # *between* the open/close tag pair - to avoid false # results where have two similar tag pairs on same line. #------------------------------ # PART 5: URLS # This is quite tricky. The URL expressions in the file # are of two kinds, of which the first is a string # between double quotes ("..") which may include # whitespace. For this case we might have a regex: url = re.compile('href=("[^">]+")', re.I) # The second case does not have quotes, and does not # allow whitespace, consisting of a continuous sequence # of non-whitespace material (that ends when you reach a # space or close bracket '>'). This might be: url = re.compile('href=([^">s]+)', re.I) # We can combine these two cases as follows, and still # get the expression back as group(1): url = re.compile(r'href=("[^">]+"|[^">s]+)', re.I) # Note that I've done nothing here to exclude 'mailto:' # links as being accepted as URLS. #------------------------------ with open('RGX_DATA.html') as infs: linenum = 0 for line in infs: linenum += 1 if line.strip() == '': continue print(' ', '-' * 100, '[%d]' % linenum, 'n TEXT:', line, end='') # PART 1: find HTML tags # (The following uses 'finditer' to find ALL matches # within the line) mm = tag.finditer(line) for m in mm: print('** TAG:', m.group(1), ' + [%s]' % m.group(2)) # PART 2,3: find open/close tags (+ params of open tags) mm = openTag.finditer(line) for m in mm: print('** OPENTAG:', m.group(1)) if m.group(2): for param in m.group(2).split(): print(' PARAM:', param) mm = closeTag.finditer(line) for m in mm: print('** CLOSETAG:', m.group(1)) # PART 4: find open/close tag pairs appearing on same line mm = openCloseTagPair.finditer(line) for m in mm: print("** PAIR [%s]: "%s"" % (m.group(1), m.group(3))) # PART 5: find URLs: mm = url.finditer(line) for m in mm: print('** URL:', m.group(1)) # PART 6: Strip out HTML tags (note that .sub will do all # possible substitutions, unless number is limited by count # keyword arg - which is fortunately what we want here) stripped = tag.sub('', line) print('** STRIPPED:', stripped, end = '')