diff --git a/textc.py b/textc.py
index 46bb44d..2e96ffe 100644
--- a/textc.py
+++ b/textc.py
@@ -14,12 +14,153 @@
 # You should have received a copy of the GNU Affero General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 from time import strftime, localtime, timezone
-from urllib.parse import unquote
+from urllib.parse import unquote, urlsplit
+from html.parser import HTMLParser
+
+
+class TGHTMLParser(HTMLParser):
+    """Only for calculate message's length and cut message.
+    Do not send these message entries to Telegram."""
+
+    def __init__(self):
+        HTMLParser.__init__(self)
+        self.messageEntries = []
+        self.s = ''
+        self.__tagL = []
+        self.__entryL = []
+
+    def close(self):
+        """This function will do some check for message entries after close."""
+        HTMLParser.close(self)
+        for e in self.messageEntries:
+            e['dataend'] = e['end']
+            e['end'] = e['end'] + len(e['tag']) + 3
+
+    def feed(self, s: str):
+        """This function will repelace \\n with &lt;br&gt;.
+        Telegram do not support &lt;br&gt; and use \\n."""
+        HTMLParser.feed(self, s.replace('\n', '<br>'))
+
+    def handle_data(self, data: str):
+        self.s = self.s + data
+        if len(self.__entryL) > 0:
+            if self.__entryL[-1]['datastart'] == -1:
+                self.__entryL[-1]['datastart'] = self.getpos()[1]
+
+    def handle_starttag(self, tag: str, attrs):
+        if len(self.__entryL) > 0:
+            if self.__entryL[-1]['datastart'] == -1:
+                self.__entryL[-1]['datastart'] = self.getpos()[1]
+        typ = ''
+        if tag in ['b', 'strong']:
+            typ = 'bold'
+        elif tag in ['i', 'em']:
+            typ = 'italic'
+        elif tag in ['u', 'ins']:
+            typ = 'underline'
+        elif tag in ['s', 'strike', 'del']:
+            typ = 'strikethrough'
+        elif tag in ['a']:
+            typ = 'url'
+            link = None
+            for name, value in attrs:
+                if name == 'href':
+                    link = value
+                    break
+        elif tag in ['code', 'pre']:
+            typ = tag
+        elif tag == 'br':
+            self.s = self.s + '\n'
+            return
+        else:
+            return  # unsuppoted tag
+        t = {'type': typ, 'offset': len(self.s), 'length': 0, 'tag': tag, 'start': self.getpos()[
+            1], 'end': 0, 'datastart': -1}
+        if typ == 'url' and link is not None:
+            t['url'] = link
+        self.__entryL.append(t)
+        self.messageEntries.append(t)
+        self.__tagL.append(tag)
+
+    def handle_endtag(self, tag: str):
+        if len(self.__entryL) > 0:
+            if self.__entryL[-1]['datastart'] == -1:
+                self.__entryL[-1]['datastart'] = self.getpos()[1]
+        if tag not in ['b', 'strong', 'i', 'em', 'u', 'ins', 's', 'strike', 'del', 'a', 'pre', 'code']:
+            return
+        if len(self.__tagL) <= 0:
+            return
+        if self.__entryL[-1]['datastart'] == -1:
+            self.__entryL[-1]['datastart'] = self.getpos()[1]
+        self.__entryL[-1]['end'] = self.getpos()[1]
+        self.__entryL[-1]['length'] = len(self.s) - self.__entryL[-1]['offset']
+        self.__entryL = self.__entryL[:-1]
+        self.__tagL = self.__tagL[:-1]
+
+
+class MessageEntries:
+    def __init__(self, l: list):
+        self.__list = l
+
+    def __len__(self):
+        return len(self.__list)
+
+    def getStr(self, l: int, s: str, re: str) -> (str, str):
+        """split string
+        s is origin string
+        re is parsed string"""
+        s = s.replace(
+            '\n', '<br>')  # the start and end is needed work with <br>
+        t = sorted(self.__list, key=lambda d: d['start'])
+        r = ''
+        i = 0
+        while i < l:
+            for v in t:
+                if v['offset'] == i:
+                    r = r + s[v['start']:v['datastart']]
+            r = r + re[i]
+            for v in reversed(t):
+                if v['offset'] + v['length'] - 1 == i:
+                    r = r + s[v['dataend']:v['end']]
+            i = i + 1
+        for v in reversed(t):
+            if i - 1 >= v['offset'] and v['offset'] + v['length'] - 1 > i - 1:
+                r = r + s[v['dataend']:v['end']]
+        r2 = ''
+        i = l
+        for v in t:
+            if i > v['offset'] and v['offset'] + v['length'] - 1 >= i:
+                r2 = r2 + s[v['start']:v['datastart']]
+        while i < len(re):
+            for v in t:
+                if v['offset'] == i:
+                    r2 = r2 + s[v['start']:v['datastart']]
+            r2 = r2 + re[i]
+            for v in reversed(t):
+                if v['offset'] + v['length'] - 1 == i:
+                    r2 = r2 + s[v['dataend']:v['end']]
+            i = i + 1
+        return r, r2
+
+    def isOkWithOrigin(self, l: int) -> bool:
+        for v in self.__list:
+            if l > v['start'] and l < v['end']:
+                return False
+        return True
+
+    def isOkWithRe(self, l: int) -> bool:
+        for v in self.__list:
+            if l > v['offset'] and l < (v['offset'] + v['length']):
+                return False
+        return True
 
 
 class textc:
     def __len__(self):
-        return len(self.__str)
+        p = TGHTMLParser()
+        p.feed(self.__str)
+        p.close()
+        return len(p.s)
 
     def __init__(self):
         self.__str = ''
@@ -32,20 +173,42 @@ class textc:
         return len(self) <= self.__max
 
     def cut(self):
-        """TODO: Need check html is not breaked
-        TODO: Need calculate char limits after parsed"""
+        """Split string"""
+        p = TGHTMLParser()
+        p.feed(self.__str)
+        p.close()
+        m = MessageEntries(p.messageEntries)
         l = self.__str.splitlines(True)
-        r = ''
-        while len(f"{r}{l[0]}") <= self.__max:
-            r = r + l[0]
-            l = l[1:]
-        if len(r) == 0:
-            r = l[0][:self.__max]
-            l[0] = l[0][self.__max:]
+        l2 = p.s.splitlines(True)
+        originlen = []  # line's origin length
+        rlen = []  # line's length after parsing
         t = ''
-        for i in l:
-            t = t + i
-        self.__str = t
+        t2 = ''
+        z = 0  # calculate the offset because of replace \n with <br>
+        for i in range(min(len(l), len(l2))):
+            if l[i].endswith('\n'):
+                z = z + 1
+            t = t + l[i]
+            t2 = t2 + l2[i]
+            originlen.append(len(t) + 3 * z)
+            rlen.append(len(t2))
+        for i in reversed(range(len(originlen))):
+            if rlen[i] <= self.__max:  # check the length
+                # make sure not break HTML
+                if m.isOkWithOrigin(originlen[i]) and m.isOkWithRe(rlen[i]):
+                    r = ''
+                    for k in range(i + 1):
+                        r = r + l[k]
+                    l = l[i+1:]
+                    t = ''
+                    for i in l:
+                        t = t + i
+                    self.__str = t
+                    return r
+                else:
+                    r, self.__str = m.getStr(rlen[i], self.__str, p.s)
+                    return r
+        r, self.__str = m.getStr(self.__max, self.__str, p.s)
         return r
 
     def tostr(self, maxLength: int = 4096):