软件翻译中，用像素判断字段超长靠谱吗？

J.H. Wang wrote:

感觉就像是用尺子量长短一样，呵呵。做得好的话，应该可以的。

Li Jie wrote:

我只是纳闷，像素怎么计算超长，同样一个汉字，"一"和"曦"的像素会不会差很多？

Li Jie,

Long time, no see. I hope everything is going well for you.

From reading what you described, my guess is that the operative word here is width. The horizontal pixel length of the two Chinese characters "一" and "嚱" are actually roughly the same. I could be wrong, but the Chinese character "了" might be the character that occupies the least horizontal pixels compared to all others.

In English, the lowercase letters "i" and "l", the exclamation mark "!", and the separator "|" obviously all occupy the least width in horizontal pixels. The uppercase "W", on the other hand, probably sits on the other end of the spectrum. But all in all, For English texts that have a certain word count, everything in total pixel length should all average out.

Anyway, it's probably a tool just to automate the irksome word-counting issue in the English and Chinese pair. Like J. H. said, it's like literally attempting to measure with a yardstick the total length of all the solid pixels, with the goal of leaving out all the white spaces.

lbone wrote:

在编程里，检测字符串长度是一项最基本的工作，有专门的函数干这活，一般来说检测的是字节数。一般来说，一个英文字母为一个字节长，一个中文字符为2个字节长，中文字符又称为宽字节或双字节字符，就是这个原因。

如果确实如你所说，要测的是字段（field）的长度的话……在软件界面上，字段一般指界面上的界面控件/界面元素，也有人用它指这个界面元素里的那个参数\变量，或它的值，在数据库和数据集里，一个数据属性或表格里的一列就是一个字段。所以说到一个字段大概可能有三种意思：
1）一个参数，或数据集里的一个属性
2）软件/网页界面里容纳这个参数的那个输入控件/界面元素或表格里的相应的列
3）这个参数或属性的值，这个第三条照理不是字段本身的定义，只是字段的值，但常有人说字段时指这个意思。

如果是（2）界面元素或表格中列的宽度等等涉及空间尺度的，这种字段（空间元素）的长度就是按像素算的。
如果是（1）或（3），它指一串字符、数字或其他类型的数据，但就算其他类型，多半还是写为一串字符串或数字的样子。这种情况，长度一般按字节算。

在界面或表格（表格也是一种界面）里的一个字符按像素算有多宽，和它的字体有关。有些字体是等宽字体，这时所有单字节字符都一样宽，所有双字节字符都是两个单字节字符那么宽。也有些字体是不等宽字体，那不一样的英文字母或汉字，就不一样宽了。

这和你这个翻译也没关系吧。我觉得翻译需要关心的主要还是按字节算的字符串的长度。

[Edited at 2014-12-24 13:56 GMT]

Li Jie wrote:

I have one more question: according to your explanation, all the white spaces will be left out. Does that mean a blank space will not be taken into account in their pixel length? Actually if counted by byte, a blank space will be considered as occupying a byte in the string. Maybe that's why I think some strings are OK but their tool reports overlength. Am I right?

Li Jie wrote:

谢谢 lbone 的详细解释。

抱歉，我开贴时用词不准确，就是字符串长度。早前做软件翻译，也都是按字节算字符串长度的。只有这个客户说用像素计算。我之所以问，是因为他们的工具生成 100 多条 truncated strings 要我改。 我随机挑了一些按字节数算了一下，并没有超长。我已经有 10 年没做软件翻译了，生怕现在有什么新方法/工具我不了解。

另外，报告中不能正确显示中文字符应该也是导致误报的原因，乱码占的字节比正常中文多很多。

Li Jie,

You're correct. Assuming that the design concept of the tool works the way we think it does, none of the horizontal whitespaces will count towards the total horizontal pixel length, effectively discounting all the blank spaces in a text.

Merry Christmas to you too.


~*~*~*~*~*

p.s., "嚱" and "曦" look to be the same character to me without going the extra step of zooming in. That's how good my eyesight is. For traditional Chinese characters, the details get a little blurry when the font is too small. But I'm still pretty happy with my current eyesight.


[Edited at 2014-12-24 19:19 GMT]

lbone wrote:

有很多小公司第一次做国际化，很多东西都还不太懂。做翻译做多了，难免会碰到一些这样不太懂的客户。
一般而言，如果你译为中文的话，字节数和长度都比英文少，所以做中文本地化或说英译中时，很少会碰到需要调整译文字数的情况。明白的开发者会在界面上留出足够的宽度，而不是让其他语言的译者去缩字。如果连中文都不够放，其他语言的麻烦就大了。
确实如steve所说，就算使用了非等宽字体（比例字体），不同字符的宽度有些差异，但字数稍多一点时，会有平均的效果，实际上最后宽度的比例关系大致还是和字节数的对比关系比较契合。客户有100多条需要你改，应该还是他们开发国际软件的经验不足，为意外情况预留的空间少了（短了）。

译文哪能随便改啊，我碰到这种情况，一般会回信强烈呼吁他们去和开发者反映，把字段的空间留大点。

[Edited at 2014-12-24 16:09 GMT]

QHE wrote:

For Microsoft Word 2007 or 2010

1. Open Microsoft Word

2. Click the Word Options button.

3. Select Advanced in the left pane.

4. Scroll down to the Display section.

5. Use the Show measurements in units of drop-down to select from Inches, Centimeters, Millimeters, Points, or Picas.

6. Click OK.

Convert between picas (computer) and pixels
http://www.unitconversion.org/typography/picas-computer-to-pixels-y-conversion.html


*** *** ***


Measuring Text (API)
http://docs.oracle.com/javase/tutorial/2d/text/measuringtext.html

Measure String Method
http://msdn.microsoft.com/en-us/library/aa327655(v=vs.71).aspx

Determine Pixel Length of String in Javascript/jQuery?
http://stackoverflow.com/questions/2057682/determine-pixel-length-of-string-in-javascript-jquery

Get the width of a string in pixels (PHP)
http://www.simplemachines.org/community/index.php?topic=122913.0





[Edited at 2014-12-24 18:14 GMT]

Li Jie wrote:

Many thanks, QHE. Merry Christmas!

QHE wrote:

Thanks, Li Jie.

Wishing you and your family a happy peaceful New Year!

[Edited at 2014-12-26 16:01 GMT]