buildz 0.6.1__tar.gz → 0.6.2__tar.gz

{buildz-0.6.1/buildz.egg-info → buildz-0.6.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: buildz
-Version: 0.6.1
+Version: 0.6.2
 Summary: 配置读写（基于json格式进行简化）、ioc、以及其他工具代码。a json-base file format's read and write code by python, and codes to read and product object from configure file in such format(ioc), and other tool codes
 Home-page: https://github.com/buildCodeZ/buildz
 Author: Zzz

buildz-0.6.2/buildz/xf/code.py

@@ -0,0 +1,71 @@
+#coding=utf-8
+
+maps = {
+    '"': '"', '\\': '\\', '/': '/',
+    'b': '\b', 'f': '\f', 'n': '\n', 'r': '\r', 't': '\t',
+}
+bmaps = {k.encode()[0]:v.encode()[0] for k,v in maps.items()}
+barr = [-1]*256
+for k,v in bmaps.items():
+    barr[k] = v
+
+pass
+def ustr(s):
+    i = 0
+    l=len(s)
+    rs = ""
+    while i<l:
+        c = s[i]
+        i+=1
+        if c == '\\':
+            x = s[i]
+            i+=1
+            if x == 'u':
+                v = int(s[i:i+4], 16)
+                r = chr(v)
+                rs+=r
+                i+=4
+                continue
+            elif x in maps:
+                r = maps[x]
+                rs+=r
+                continue
+        rs+=c
+    return rs
+
+pass
+bl2,bu = b'\\u'
+def ubytes(s, code = "utf-8"):
+    i = 0
+    l=len(s)
+    rs = b""
+    rs = []
+    while i<l:
+        c = s[i]
+        i+=1
+        if c != bl2:
+            rs.append(c)
+            continue
+        x = s[i]
+        i+=1
+        rx = barr[x]
+        if rx>=0:
+            rs.append(rx)
+            continue
+        elif x == bu:
+            v = int(s[i:i+4], 16)
+            r = chr(v).encode(code)
+            rs+=list(r)
+            i+=4
+            continue
+        rs.append(c)
+        rs.append(x)
+    #print(rs)
+    rs = bytes(rs)
+    return rs
+
+pass
+def ub2s(s, code = "utf-8"):
+    return ubytes(s, code).decode(code)
+
+pass

buildz-0.6.2/buildz/xf/code_modify.py

@@ -0,0 +1,71 @@
+#coding=utf-8
+
+maps = {
+    '"': '"', '\\': '\\', '/': '/',
+    'b': '\b', 'f': '\f', 'n': '\n', 'r': '\r', 't': '\t',
+}
+bmaps = {k.encode()[0]:v.encode()[0] for k,v in maps.items()}
+barr = [-1]*256
+for k,v in bmaps.items():
+    barr[k] = v
+
+pass
+def ustr(s):
+    i = 0
+    l=len(s)
+    rs = ""
+    while i<l:
+        c = s[i]
+        i+=1
+        if c == '\\':
+            x = s[i]
+            i+=1
+            if x == 'u':
+                v = int(s[i:i+4], 16)
+                r = chr(v)
+                rs+=r
+                i+=4
+                continue
+            elif x in maps:
+                r = maps[x]
+                rs+=r
+                continue
+        rs+=c
+    return rs
+
+pass
+bl2,bu = b'\\u'
+def ubytes(s, code = "utf-8"):
+    i = 0
+    l=len(s)
+    rs = b""
+    rs = []
+    while i<l:
+        c = s[i]
+        i+=1
+        if c != bl2:
+            rs.append(c)
+            continue
+        x = s[i]
+        i+=1
+        rx = barr[x]
+        if rx>=0:
+            rs.append(rx)
+            continue
+        elif x == bu:
+            v = int(s[i:i+4], 16)
+            r = chr(v).encode(code)
+            rs+=list(r)
+            i+=4
+            continue
+        rs.append(c)
+        rs.append(x)
+    #print(rs)
+    rs = bytes(rs)
+    return rs
+
+pass
+def ub2s(s, code = "utf-8"):
+    return ubytes(s, code).decode(code)
+
+pass

buildz-0.6.2/buildz/xf/loaderz/deal/strz.py

@@ -0,0 +1,284 @@
+from .. import base
+from .. import item
+from .. import exp
+from ... import file
+from ... import code as codez
+from . import lr
+
+def init():
+    cs = "abfnrtv\\'\"?0123456789xcde"
+    maps = {k:k.encode()[0] for k in cs}
+    global c2b
+    c2b = maps
+    global special_keys
+    special_keys = [-1]*256
+    s  = "abfnrtv\\'\""
+    ts = b"\a\b\f\n\r\t\v\\\'\""
+    for c,t in zip(s, ts):
+        special_keys[c2b[c]] = t
+    symbal_a = [False]*256
+    for c in s:
+        symbal_a[c2b[c]] = True
+    global id_0,id_a,id_x,id_A
+    global c2nums
+    c2nums = [-1]*256
+    id_0,id_a,id_A,id_x,id_u = b'0aAxu'
+    for i in range(128):
+        v = i-id_0
+        if v>=0 and v<=9:
+            c2nums[i] = v
+            continue
+        v = i-id_a
+        if v>=0 and v<=5:
+            c2nums[i]=v+10
+            continue
+        v = i-id_A
+        if v>=0 and v<=5:
+            c2nums[i]=v+10
+
+pass
+init()
+def cs2num(bts, base, min, max, oct = False):
+    cnt = 0
+    rst = 0
+    mv = 4
+    if oct:
+        mv = 3
+    for i in range(max):
+        if base+i>=len(bts):
+            if i<min:
+                return -1,0
+            else:
+                break
+        bt = bts[base+i]
+        v = c2nums[bt]
+        if v<0 or (oct and v>9):
+            if i<min:
+                return -1,0
+            else:
+                break
+        cnt+=1
+        rst = (rst<<mv)|v
+    return rst, cnt
+
+pass
+def decode_u(val):
+    """
+    1 byte: 7 0? 
+    2 byte: 11 110? 10?
+    3 byte: 16 1110? 10? 10?
+    4 byte: 21 11110? 10? 10? 10?
+    """
+    if val<0x80:
+        return [val]
+    elif val<0x800:
+        b0 = 0x80|(val&0x7f)
+        b1 = 0x60|(val>>6)
+        return [b1, b0]
+    elif val<0x10000:
+        b0 = 0x80|(val&0x3f)
+        val>>=6
+        b1 = 0x80|(val&0x3f)
+        b2 = 0xe0|(val>>6)
+        return [b2,b1,b0]
+    else:
+        b0 = 0x80|(val&0x3f)
+        val>>=6
+        b1 = 0x80|(val&0x3f)
+        val>>=6
+        b2 = 0x80|(val&0x3f)
+        b3 = 0xf0|(val>>6)
+        return [b3,b2,b1,b0]
+
+pass
+def translate_bts(bts, octs = None, hexs = None):
+    i = 0
+    rs = []
+    while i<len(bts):
+        c = bts[i]
+        i+=1
+        if c!=c2b['\\']:
+            rs.append(c)
+            continue
+        x0 = bts[i]
+        i+=1
+        if special_keys[x0]>=0:
+            rs.append(special_keys[x0])
+            continue
+        if x0==b'u'[0]:
+            c_val, c_cnt = cs2num(bts, i, 4,4, 0)
+            if c_cnt==0:
+                raise Exception("\\uXXXX error")
+            tmp = decode_u(c_val)
+            i+=c_cnt
+            rs+=tmp
+            continue
+        if x0==id_x:
+            c_val, c_cnt = cs2num(bts, i, 2, 2, 0)
+            if c_cnt==0:
+                raise Exception("\\xXX error")
+            if hexs is not None:
+                tmp = hexs[c_val]
+            else:
+                tmp = [c_val]
+            rs+=tmp
+            i+=c_cnt
+            continue
+        c_val,c_cnt = cs2num(bts, i-1, 1, 3, 1)
+        if c_cnt>0:
+            if octs is not None:
+                tmp = octs[c_val]
+            else:
+                tmp = [c_val%256]
+            rs+=tmp
+            i+=c_cnt-1
+            continue
+        rs.append(c)
+        rs.append(x0)
+        #i-=1
+    return bytes(rs)
+
+pass
+def gen_chars(code="utf-8"):
+    simple = "abfnrtv\\'\""
+    octs = [0]*512
+    hexs = [0]*256
+    for i in range(512):
+        if i<256:
+            vhex = hex(i)[2:]
+            if len(vhex)==1:
+                vhex = "0"+vhex
+            cmd = f"'\\x{vhex}'"
+            hexs[i] = list(eval(cmd).encode(code))
+        voct = oct(i)[2:]
+        cmd = f"'\\{voct}'"
+        octs[i] = list(eval(cmd).encode(code))
+    return octs, hexs
+
+pass
+class PrevStrDeal(lr.LRDeal):
+    def types(self):
+        if not self.deal_build:
+            return []
+        return ['str']
+    def build(self, obj):
+        obj.is_val = 1
+        if self.translate:
+            val = obj.val
+            val = self.do_translate(val)
+            obj.val = val
+        return obj
+    def prepare(self, mg):
+        super().prepare(mg)
+        self.as_bytes = mg.as_bytes
+        #if not self.as_bytes:
+        #    self.octs,self.hexs = gen_chars()
+        self.label_l2 = mg.like("\\")
+        self.label_qt = mg.like('"')
+        self.label_et = mg.like("\n")
+        self.label_lr = mg.like("\r")
+        self.label_nl = mg.like("")
+        self.et_in_right = self.right.count(self.label_et)
+    def init(self, left = '"', right= '"', single_line = False, note = False, translate = False, deal_build = False):
+        super().init(left, right, 'str')
+        self.single_line = single_line
+        self.note = note
+        self.translate = translate
+        self.deal_build = deal_build
+        self.as_bytes = True
+        self.octs = None
+        self.hexs = None
+    def json_loads(self, s):
+        import json
+        x = s
+        cd = None
+        if type(x)==bytes:
+            x, cd = file.decode_c(x)
+        rs = json.loads(x)
+        if type(s)==bytes:
+            rs = rs.encode(cd)
+        return rs
+    def do_translate(self, s):
+        """
+            取巧用python的eval来生成字符表
+        """
+        is_bytes = type(s)==bytes
+        if is_bytes:
+            return codez.ubytes(s, "utf-8")
+        else:
+            return codez.ub2s(s.encode("utf-8"), "utf-8")
+            return codez.ustr(s)
+        if not is_bytes:
+            s = s.encode("utf-8")
+        #s = s.decode("unicode_escape")
+        s = translate_bts(s, self.octs, self.hexs)
+        if is_bytes:
+            #s = s.encode("utf-8")
+            s = s.decode("utf-8")
+        return s
+        """
+            取巧直接调用json
+        """
+        qt = self.label_qt
+        ql = self.label_l2
+        et = self.label_et
+        tr = self.label_lr
+        nt = self.label_nl
+        pt = ql+qt
+        arr = s.split(pt)
+        arr = [k.replace(qt, pt) for k in arr]
+        s = pt.join(arr)
+        #s = s.replace(qt, ql+qt)
+        s = s.replace(tr, nt)
+        arr = s.split(et)
+        outs = [self.json_loads(qt+k+qt) for k in arr]
+        outs = et.join(outs)
+        return outs
+    def deal(self, buffer, rst, mg):
+        cl = buffer.read(self.ll)
+        if cl != self.left:
+            return False
+        rm = buffer.full().strip()
+        buffer.clean2read(self.ll)
+        if len(rm)>0:
+            if not self.note:
+                raise Exception(f"unexcept char before string: {rm}")
+            else:
+                rst.append(item.Item(rm, type = "str", is_val = 0))
+        tmp = cl[:0]
+        ctmp = tmp[:0]
+        do_judge = 1
+        mark_et = 0
+        mark_l2 = 0
+        while True:
+            if do_judge and self.right == buffer.rget(self.lr):
+                break
+            c = buffer.read_cache(1)
+            if do_judge and c == self.label_et:
+                mark_et += 1
+            if len(c)==0:
+                if self.single_line and self.note:
+                    break
+                raise Exception(f"unexcept string end while reading str")
+            do_judge = 1
+            if c == self.label_l2:
+                mark_l2 = 1
+                do_judge = 0
+                c = buffer.read_cache(1)
+                if len(c)==0:
+                    raise Exception(f"unexcept string end while reading str")
+        data = buffer.full()
+        data = data[:-self.lr]
+        buffer.clean()
+        mark_et -= self.et_in_right
+        if self.single_line and mark_et>0:
+            print("left:",self.left, "right:", self.right)
+            raise Exception(f"contain enter in single line string")
+        if self.translate and mark_l2:
+            data = self.do_translate(data)
+        if self.note:
+            return True
+        rst.append(item.Item(data, type='str', is_val = 1))
+        return True
+
+pass

buildz-0.6.1/buildz/xf/loaderz/deal/strz.py → buildz-0.6.2/buildz/xf/loaderz/deal/strz_new.py

@@ -206,6 +206,7 @@ class PrevStrDeal(lr.LRDeal):
         if is_bytes:
             return codez.ubytes(s, "utf-8")
         else:
+            return codez.ub2s(s.encode("utf-8"), "utf-8")
             return codez.ustr(s)
         if not is_bytes:
             s = s.encode("utf-8")

buildz-0.6.2/buildz/xf/loaderz/deal/strz_old.py

@@ -0,0 +1,197 @@
+from .. import base
+from .. import item
+from .. import exp
+from ... import file
+import json
+from . import lr
+from ... import code as codez
+def init():
+    cs = "abfnrtv\\'\"?0123456789xcde"
+    maps = {k:k.encode()[0] for k in cs}
+    global c2b
+    c2b = maps
+    global symbal_a
+    global a_map
+    a_map = [0]*256
+    s  = "abfnrtv\\'\"?"
+    ts = b"\a\b\f\n\r\t\v\\\'\"\?"
+    for c,t in zip(s, ts):
+        a_map[c2b[c]] = t
+    symbal_a = [False]*256
+    for c in s:
+        symbal_a[c2b[c]] = True
+    global id_0,id_a,id_x
+    id_0 = b'0'[0]
+    id_a = b'a'[0]
+    id_x = b'x'[0]
+
+pass
+
+init()
+def translate_bts(bts):
+    i = 0
+    rs = []
+    while i<len(bts):
+        c = bts[i]
+        i+=1
+        if c!=c2b['\\']:
+            rs.append(c)
+            continue
+        x0 = bts[i]
+        i+=1
+        if symbal_a[x0]:
+            rs.append(a_map[x0])
+            continue
+        v = x0-id_0
+        if v>=0 and v<=7:
+            tmp = v
+            for j in range(2):
+                if i+j+1>=len(bts):
+                    break
+                xi = bts[i+j+1]
+                vi = xi-id_0
+                if vi>=0 and vi<=7:
+                    i+=1
+                    tmp = (rs<<3)+vi
+                else:
+                    break
+            rs.append(tmp)
+            continue
+        if x0 == id_x:
+            tmp = 0
+            for j in range(2):
+                if i+j+1>=len(bts):
+                    raise Exception("\\xXX error")
+                xi = bts[i+j+1]
+                vi = xi-id_0
+                if vi<0 or vi>7:
+                    vi = xi-id_a
+                    if vi<0 or vi > 5:
+                        raise Exception("\\xXX error")
+                    vi +=10
+                tmp=(tmp<<4)+vi
+            rs.append(tmp)
+            continue
+        rs.append(c)
+        rs.append(x0)
+        #i-=1
+    return bytes(rs)
+
+pass
+
+class PrevStrDeal(lr.LRDeal):
+    def types(self):
+        if not self.deal_build:
+            return []
+        return ['str']
+    def build(self, obj):
+        if type(obj.val)==list:
+            return None
+        if obj.is_val:
+            print("error str build")
+            return obj
+        obj.is_val = 1
+        if self.translate:
+            print("error str build tr")
+            val = obj.val
+            val = self.do_translate(val)
+            obj.val = val
+        return obj
+    def prepare(self, mg):
+        super().prepare(mg)
+        self.label_l2 = mg.like("\\")
+        self.label_qt = mg.like('"')
+        self.label_et = mg.like("\n")
+        self.label_lr = mg.like("\r")
+        self.label_nl = mg.like("")
+        self.et_in_right = self.right.count(self.label_et)
+    def init(self, left = '"', right= '"', single_line = False, note = False, translate = False, deal_build = False):
+        super().init(left, right, 'str')
+        self.single_line = single_line
+        self.note = note
+        self.translate = translate
+        self.deal_build = deal_build
+    def json_loads(self, s):
+        x = s
+        cd = None
+        if type(x)==bytes:
+            x, cd = file.decode_c(x)
+        rs = json.loads(x)
+        if type(s)==bytes:
+            rs = rs.encode(cd)
+        return rs
+    def do_translate(self, s):
+        is_bytes = type(s)==bytes
+        if not is_bytes:
+            s = s.encode("utf-8")
+        #s = translate_bts(s)
+        s = codez.ubytes(s, "utf-8")
+        if not is_bytes:
+            s = s.decode("utf-8")
+        return s
+        """
+            取巧直接调用json
+        """
+        qt = self.label_qt
+        ql = self.label_l2
+        et = self.label_et
+        tr = self.label_lr
+        nt = self.label_nl
+        pt = ql+qt
+        arr = s.split(pt)
+        arr = [k.replace(qt, pt) for k in arr]
+        s = pt.join(arr)
+        #s = s.replace(qt, ql+qt)
+        s = s.replace(tr, nt)
+        arr = s.split(et)
+        outs = [self.json_loads(qt+k+qt) for k in arr]
+        outs = et.join(outs)
+        return outs
+    def deal(self, buffer, rst, mg):
+        cl = buffer.read(self.ll)
+        if cl != self.left:
+            return False
+        rm = buffer.full().strip()
+        buffer.clean2read(self.ll)
+        if len(rm)>0:
+            if not self.note:
+                raise Exception(f"unexcept char before string: {rm}")
+            else:
+                rst.append(item.Item(rm, type = "str", is_val = 0))
+        tmp = cl[:0]
+        ctmp = tmp[:0]
+        do_judge = 1
+        mark_et = 0
+        mark_l2 = 0
+        while True:
+            if do_judge and self.right == buffer.rget(self.lr):
+                break
+            c = buffer.read_cache(1)
+            if do_judge and c == self.label_et:
+                mark_et += 1
+            if len(c)==0:
+                if self.single_line and self.note:
+                    break
+                raise Exception(f"unexcept string end while reading str")
+            do_judge = 1
+            if c == self.label_l2:
+                mark_l2 = 1
+                do_judge = 0
+                c = buffer.read_cache(1)
+                if len(c)==0:
+                    raise Exception(f"unexcept string end while reading str")
+        data = buffer.full()
+        data = data[:-self.lr]
+        buffer.clean()
+        mark_et -= self.et_in_right
+        if self.single_line and mark_et>0:
+            print("left:",self.left, "right:", self.right)
+            raise Exception(f"contain enter in single line string")
+        if self.translate and mark_l2:
+            data = self.do_translate(data)
+        if self.note:
+            return True
+        rst.append(item.Item(data, type='str', is_val = 1))
+        return True
+
+pass

{buildz-0.6.1 → buildz-0.6.2}/buildz/xf/loaderz/test.py

@@ -12,19 +12,20 @@ xf
 time cost in <function loads at 0x0000015DEDD0F100>: 0.3307473659515381
 
 慢7倍多，感觉还能接受，之前没注意速度，更早的版本(buildz.xf.read.loads，后面可能会删掉)耗时要三四秒，对比之后就重新写了个，也不管什么堆栈了，就递归调用，后面有时间再改回堆栈吧（python的list的append和pop效率貌似不咋地，尤其是pop(0)）
-又写了个C版本的代码来加速，速度大概比C加速版json的慢三到四倍（汗，和没用C加速版的json差不多速度，追求结构化不追求速度，用C只是因为C本身比python快，起码满足自己使用了，本测试代码里C加速版xf比python版快7倍，在其他场景里实际使用快了近一百倍），源码暂不公布（C语言要自己写垃圾回收真累，还要自己写List和Map，另外就是用C写了后，发现除了当库给其他语言使用，貌似没啥用）
+又写了个C版本的代码来加速，速度大概比C加速版json的慢三到四倍（汗，和没用C加速版的json差不多速度，追求结构化不追求速度，逻辑上本来就多了不少内存分配和运算，用C只是因为C本身比python快，起码满足自己使用了，本测试代码里C加速版xf比python版快7倍，在其他场景里实际使用快了十倍以上，不过也比C版json慢了六七倍），源码暂不公布（C语言要自己写垃圾回收真累，还只是写了个计数器的，还要自己写List和Map，另外就是用C写了后，发现除了当库给其他语言使用，貌似没啥用，还有cython调用C真方便，后续可能会优化下，配置文件格式错误把错误位置打印出来，这个实现起来简单，不过不一定写，不是必要的，还有C版本的垃圾回收在抛错的时候清理下数据，不清理的话，配置文件格式错误会导致之前分配的内存未释放，内存占用增多，因为不打算直接在C用xf，打算抛错的时候，C用的内存全释放算了，之后调用再重新分配）
+还有个比较操蛋的点是，用java也写了一版读取xf格式配置文件，结果java版比C版还快，一脸问号，是在C写的内存分配和回收占用时间了吗？更无语的是，java做循环读取配置文件，貌似会触发java的优化，平均速度还会更快（循环100次后和C加速版python的json差不多一样快了），不清楚是因为循环读的是一样的字符串导致的java优化，还是其他原因导致的java优化，如果是其他原因，说明读多个配置文件的时候java版会越来越快，java版会比C版更适合使用，而且java自己有垃圾回收，有List和Map，有时间下个python库用python调java试试
 """
 from buildz.xf import readz as rz
 from buildz.xf import read as rd
 from buildz import xf, fz
 import json
-import time
+import time,sys
 
 try:
     # C加速代码
     import cxf
 except:
-    from buildz.xf import read as cxf
+    from buildz.xf import readz as cxf
     pass
 pass
 def cost(n, f,*a,**b):

{buildz-0.6.1 → buildz-0.6.2}/buildz/xf/readz.py

@@ -43,7 +43,11 @@ def build_val(mgs):
     mgs.add(reval.ValDeal("false", lambda x:False))
 
 pass
+g_mg = None
 def build(as_bytes=False):
+    global g_mg
+    if g_mg is not None and g_mg.as_bytes == as_bytes:
+        return g_mg
     mgs = mg.Manager(as_bytes)
     mgs.add(setz.SetDeal(':'))
     mgs.add(setz.SetDeal('='))
@@ -70,6 +74,7 @@ def build(as_bytes=False):
     mgs.add(strz.PrevStrDeal("'","'",1,0,1))
     mgs.add(strz.PrevStrDeal('"','"',1,0,1))
     mgs.add(nextz.PrevNextDeal())
+    g_mg = mgs
     return mgs
 
 pass

{buildz-0.6.1 → buildz-0.6.2/buildz.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: buildz
-Version: 0.6.1
+Version: 0.6.2
 Summary: 配置读写（基于json格式进行简化）、ioc、以及其他工具代码。a json-base file format's read and write code by python, and codes to read and product object from configure file in such format(ioc), and other tool codes
 Home-page: https://github.com/buildCodeZ/buildz
 Author: Zzz