Editorial Workflows

#coding: utf-8
import workflow

TYP_ROOT = 0
TYP_HEAD = 1
TYP_UNORDERED = 2
TYP_ORDERED = 3
TYP_BODY = 4
TYP_QUOTE = 5
TYP_CODE = 6
TYP_RULE = 7
TYP_EMPTY = 8
TYP_OTHER = 9

STR_HASH = '#'
STR_STAR = '*'
STR_DASH = '-'
STR_QUOTE = '>'
STR_SPACE = ' '
STR_DOT = '.'
STR_TAB = '\t'

def main():
    lst_Nest = addParentChild(parse_text(workflow.get_input()))
    workflow.set_output(make_tabbed(lst_Nest))
    
def make_tabbed(lst_Nodes):
    str_out = ''
    for dct_Node in lst_Nodes[1:]:
        if (dct_Node['type'] in set([TYP_HEAD, TYP_ORDERED, TYP_UNORDERED])):
            n_Tabs = dct_Node['level']-1
        else:
            n_Tabs = dct_Node['level']
        str_out = ''.join([str_out, '\t' * n_Tabs, dct_Node['text']])
    return str_out
    
def addParentChild(lst_Lines):
    """Add ParentID and ChildIDs to each node"""
    # The parent of Level 1 headers (and pre-header full-left lines) is the virtual root.
    lst_Parents = [0]
    lng_CurrentHash = 0
    for dct_Line in lst_Lines:
        dct_Line['chiln'] = []
        lng_ID = dct_Line['id']
        vType = dct_Line['type']
        
        if (vType != TYP_ROOT):
            lng_Depth = dct_Line['indents'] + 1
           
            # get nesting level (prevailing header level + any tab indents)
            if (vType != TYP_HEAD):
                if (vType != TYP_EMPTY):
                    lng_Level = lng_CurrentHash + lng_Depth
                else:
                    lng_Level = 1
            else:
                lng_CurrentHash = lng_Depth
                lng_Level = lng_Depth
                
            # FIND current parent of this level
            dct_Line['level'] = lng_Level
            while (len(lst_Parents) < lng_Level):
                lst_Parents.append(lst_Parents[-1])
            id_Parent = lst_Parents[lng_Level-1] 
            
            # RECORD parent-child relationship
            dct_Line['parentID'] = id_Parent
            lst_Lines[id_Parent]['chiln'].append(lng_ID)
            
            # UPDATE current level-parent list
            if (len(lst_Parents) <= (lng_Level)):
                lst_Parents.append(lng_ID)
            else:
                lst_Parents[lng_Level] = lng_ID
    return lst_Lines
        
def parse_text(str_Text):
    """Return a list of line records
        Line properties:
            Markdown type, prefix depth (hashes/tabs), line number as id, StartPosn
    """
    lst_Lines = [{'type':TYP_ROOT, 'indents':0, 'id':0, 'text':'VIRTUAL_ROOT'}]
    lng_Posn = 0
    for i_Line, strLine in enumerate(iter(str_Text.splitlines(True))):
        lst_Lines.append(parse_line(i_Line+1, lng_Posn, strLine))
        lng_Posn += len(strLine)
    return lst_Lines
            
def parse_line(int_ID, int_Posn, str_Line):
    """
    Get the type, hash or tab level, and non-prefix text of an MD or tab-indented line
    NB code blocks are assumed start with 4 spaces. All tabbed lines are read as nested.
    """
    lng_Level = 0
    lng_Type = TYP_EMPTY
    str_Prev = ''
    lng_Space = 0
    bln_Bullet = 0
    if (str_Line != ''):
        for i, c in enumerate(str_Line):
            # FIRST CHAR OF THE LINE ?
            if (str_Prev == ''):
                if (c == STR_TAB):
                    lng_Level = 1
                elif (c == STR_DASH or c == STR_STAR):
                    bln_Bullet = 1
                elif (c == STR_SPACE):
                    lng_Space = 1
                else:
                    if (c != STR_HASH and c != STR_QUOTE and not c.isdigit()):
                        return {'id':int_ID, 'indents':0, 'type':TYP_BODY, 'posn':int_Posn, 'text':str_Line}
            # FOLLOWING A HASH ?
            elif (str_Prev == STR_HASH):
                if (c == STR_HASH):
                    lng_Level += 1
                elif (c == STR_SPACE):
                    return {'id':int_ID, 'indents':lng_Level, 'type':TYP_HEAD, 'posn':int_Posn, 'text':str_Line[i+1:]}
                else:
                    return {'id':int_ID, 'indents':0, 'type':TYP_OTHER, 'posn':int_Posn, 'text':str_Line}
            # FOLLOWING A BULLET ?
            elif (str_Prev == STR_DASH or str_Prev == STR_STAR):
                if (c == STR_DASH or c == STR_STAR):
                    bln_Bullet += 1
                    if (bln_Bullet > 2):
                        return {'id':int_ID, 'indents':0, 'type':TYP_RULE, 'posn':int_Posn, 'text':str_Line.lstrip()}
                elif (c != STR_SPACE):
                    return {'id':int_ID, 'indents':0, 'type':TYP_OTHER, 'posn':int_Posn, 'text':str_Line}
            # FOLLOWING A TAB ?
            elif (str_Prev == STR_TAB):
                if (c == STR_TAB):
                    lng_Level += 1
                elif (c == STR_DASH or c == STR_STAR):
                    bln_Bullet += 1
                elif (not c.isdigit()):
                    return {'id':int_ID, 'indents':lng_Level, 'type':TYP_BODY, 'posn':int_Posn, 'text':str_Line[i:]}
            # FOLLOWING A DIGIT ?
            elif (str_Prev.isdigit()):
                if (c != STR_DOT and not c.isdigit()):
                    return {'id':int_ID, 'indents':0, 'type':TYP_OTHER, 'posn':int_Posn, 'text':str_Line}
            # FOLLOWING A DOT ?
            elif (str_Prev == STR_DOT):
                if (c == STR_SPACE):
                    return {'id':int_ID, 'indents':lng_Level, 'type':TYP_ORDERED, 'posn':int_Posn, 'text':str_Line[i+1:]}
                else:
                    return {'id':int_ID, 'indents':0, 'type':TYP_OTHER, 'posn':int_Posn, 'text':str_Line}
            # FOLLOWING A '>' ?
            elif (str_Prev == STR_QUOTE):
                if (c == STR_SPACE):
                    return {'id':int_ID, 'indents':0, 'type':TYP_QUOTE, 'posn':int_Posn, 'text':str_Line[i+1:]}
                elif (c != STR_QUOTE) :
                    return {'id':int_ID, 'indents':0, 'type':TYP_OTHER, 'posn':int_Posn, 'text':str_Line}
            # FOLLOWING A SPACE ?
            elif (str_Prev == STR_SPACE):
                if (c == STR_SPACE):
                    lng_Space+=1
                    if (lng_Space > 3):
                        return {'id':int_ID, 'indents':0, 'type':TYP_CODE, 'posn':int_Posn, 'text':str_Line[i+1:]}
                elif (c == STR_DASH or c == STR_STAR):
                    bln_Bullet += 1
                    if (bln_Bullet > 2):
                        return {'id':int_ID, 'indents':0, 'type':TYP_RULE, 'posn':int_Posn, 'text':str_Line.lstrip()}
                elif (bln_Bullet > 0):
                    return {'id':int_ID, 'indents':lng_Level, 'type':TYP_UNORDERED, 'posn':int_Posn, 'text':str_Line[i:]}
                else:
                    return {'id':int_ID, 'indents':0, 'type':TYP_BODY, 'posn':int_Posn, 'text':str_Line}
            else:
                return {'id':int_ID, 'indents':0, 'type':TYP_OTHER, 'posn':int_Posn, 'text':str_Line}
            
            str_Prev = c
    else:
        return {'id':int_ID, 'indents':0, 'type':TYP_EMPTY, 'posn':int_Posn, 'text':''}

    return {'id':int_ID, 'indents':0, 'type':TYP_BODY, 'posn':int_Posn, 'text':str_Line}
    
main()