May « 2010 « Katz Got Your Tongue?

Yehuda Katz is a member of the Ruby on Rails core team, and lead developer of the Merb project. He is a member of the jQuery Core Team, and a core contributor to DataMapper. He contributes to many open source projects, like Rubinius and Johnson, and works on some he created himself, like Thor.

@garybernhardt @steveklabnik I guess I have to ask… is cls a troll project?

Archive for May, 2010

Encodings, Unabridged

Monday, May 17th, 2010

I wrote somewhat extensively about the problem of encodings in Ruby 1.9 in general last week.

For those who didn’t read that post, let me start with a quick refresher.

What’s an Encoding?

An encoding specifies how to take a list of characters (such as “hello”) and persist them onto disk as a sequence of bytes. You’re probably familiar with the ASCII encoding, which specifies how to store English characters in a single byte each (taking up the space in 0-127, leaving 128-255 empty).

Another common encoding is ISO-8859-1 (or Latin-1), which uses ASCII’s designation for the first 127 characters, and designates the numbers 128-255 for Latin characters (such as “é” or “ü”).

Obviously, 255 characters isn’t enough for all languages, so there are a number of ISO-8859-* encodings which each designate numbers 128-255 for their own purposes (for instance, ISO-8859-5 uses that space for Russian characters).

Unfortunately, the raw bytes themselves do not contain an “encoding specifier” or any kind, and the exact same bytes can either mean something in Western characters, Russian, Japanese, or any other language, depending on the character set that was originally used to store off the characters as bytes.

As a general rule, protocols (such as HTTP), provide a mechanism for specifying the encoding. For instance, in HTTP, you can specify the encoding in the Content-Type header, like this: Content-Type: text/html; charset=UTF-8. However, this is not a requirement, so it is possible to receive some content over HTTP and not know its encoding.

This brings us to an important point: Strings have no inherent encoding. By default, Strings are just BINARY data. Since the data
could be encoded using any number of different incompatible encodings, simply combining BINARY data from different sources could easily result in a corrupted String.

When you see a diamond with a question mark inside on the web, or gibberish characters (like a weird A with a 3/4 symbol), you’re seeing a mistaken attempt to combine binary data encoded differently into a single String.

What’s Unicode

Unicode is an effort to map every known character (to a point) to a number. Unicode does not define an encoding (how to represent those numbers in bytes). It simply provides a unique number for each known character.

Unicode tries to unify characters from different encodings that represent the same character. For instance, the A in ASCII, the A in
ISO-8859-1, and the A in the Japanese encoding SHIFT-JIS all map to the same Unicode character.

Unicode also takes pains to ensure round-tripping between existing encodings and Unicode. Theoretically, this should mean that it’s possible to take some data
encoded using any known encoding, use Unicode tables to map the characters to Unicode numbers, and then use the reverse versions of those tables to map the Unicode numbers back into the original encoding.

Unfortunately, both of these characteristics cause some problems for Asian character sets. First, there have been some historical errors in the process of
unification, which requires the Unicode committee to properly identify which characters in different existing Chinese, Japanese and Korean (CJK) character sets actually represent the same character.

In Japanese, personal names use slight variants of the non-personal-name version of the same character. This would be equivalent to the difference (in English) between “Cate” and “Kate”. Many of these characters (sometimes called Gaiji) cannot be represented in Unicode at all.

Second, there are still hundreds of characters in some Japanese
encodings (such as the Microsoft encoding to SHIFT-JIS called CP932 or Windows-31J) that simply do not round-trip through Unicode.

To make matters worse, Java and MySQL use a different mapping table than the standard Unicode mapping tables (making “This costs ¥5″ come out in Unicode as “This costs \5″). The standard Unicode mapping tables handle this particular case correctly (but cannot fully solve the round-tripping problem), but these quirks only serve to further raise doubts about Unicode in the minds of Japanese developers.

For a lot more information on these issues, check out the XML Japanese Profile document created by the W3C to explain how to deal with some of these problems in XML documents.

In the Western world, all encodings in use do not have these problems. For instance, it is trivial to take a String encoded as ISO-8859-1, convert it into
Unicode, and then convert it back into ISO-8859-1 when needed.

This means that for most of the Western world, it is a good idea to
use Unicode as the “one true character set” inside programming languages. This means that programmers can treat Strings as simple sequences of Unicode code points (several
code points may add up to a single character, such as the ¨ code point, which can be
applied to other code points to form characters like ü). In the Asian world, while this can sometimes be a good strategy, it is often significantly simpler to use the original encoding and handle merging Strings in different encodings together manually (when an appropriate decision about the tradeoffs around fidelity can be made).

Before I continue, I would note that the above is a vast simplification of the
issues surrounding Unicode and Japanese. I believe it to be a fair characterization,
but die-hard Unicode folks, and die-hard anti-Unicode folks would possibly disagree
with some elements of it. If I have made any factual errors, please let me know.

A Digression: UTF-*

Until now, I have talked only about “Unicode”, which simply maps code points
to numbers. Because Unicode uses counting numbers, it can accommodate as many
code points as it wants.

However, it is not an encoding. In other words, it does not specify how to
store the numbers on disk. The most obvious solution would be to use a few
bytes for each character. This is the solution that UTF-32 uses, specifying
that each Unicode character be stored as four bytes (accommodating over 4 billion characters). While this has the advantage of being simple, it also uses huge amounts of memory and disk space compared to the original encodings (like ASCII, ISO-8859-1 and SHIFT-JIS) that it is replacing.

On the other side of the spectrum is UTF-8. UTF-8 uses a single byte for English characters, using the exact same mapping as ASCII. This means that a UTF-8 string that contains only characters found in ASCII will have the identical bytes as a String stored in ASCII.

It then uses the high bit (the bytes representing 128-255) to specify a series of escape characters that can specify a multibyte character. This means that Strings using Western characters use relatively few bytes (often comparable with the original encodings Unicode replaces), because they are in the low area of the Unicode space, while the large number of characters in the Asian languages use more bytes than their native encodings, because they use characters with larger Unicode numbers.

This is another reason some Asian developers resent Unicode; while it does not significantly increase the memory requirements for most Western documents, it does so for Asian documents.

For the curious, UTF-16 uses 16-bits for the most common characters (the BMP, or basic multilingual plane), and 32-bits to represent characters from planes 1 through 16. This means that UTF-8 is most efficient for Strings containing mostly ASCII characters. UTF-8 and UTF-16 are approximately equivalent for Strings containing mostly characters outside ASCII but inside the the BMP. For Strings containing mostly characters outside the BMP, UTF-8, UTF-16, and UTF-32 are approximately equivalent. Note that when I say “approximately equivalent”, I’m not saying that they’re exactly the same, just that the differences are small in large Strings.

Of the Unicode encodings, only UTF-8 is compatible with ASCII. By this I mean that if a String is valid ASCII, it is also valid UTF-8. UTF-16 and UTF-32 encode ASCII characters using two or four bytes.

What Ruby 1.9 Does

Accepting that there are two very different ways of handling this problem, Ruby 1.9 has a String API that is somewhat different from most other languages, mostly influenced by the issues I described above in dealing with Japanese in Unicode.

First, Ruby does not mandate that all Strings be stored in a single internal encoding. Unfortunately, this is not possible to do reliably with common Japanese encodings (CP932, aka Windows-31J has 300 characters than cannot round-trip through Unicode without corrupting data). It is possible that the Unicode committee will some day fully solve these problems to everyone’s satisfaction, but that day has not yet come.

Instead, Ruby 1.9 stores Strings as the original sequence of bytes, but allows a String to be tagged with its encoding. It then provides a rich API for converting Strings from one encoding to another.

string = "hello"                     # by default, string is encoded as "ASCII"
string.force_encoding("ISO-8859-1")  # this simply retags the String as ISO-8859-1
                                     # this will work since ISO-8859-1
                                     # is a superset of ASCII.
 
string.encode("UTF-8")               # this will ask Ruby to convert the bytes in
                                     # the current encoding to bytes in
                                     # the target encoding, and retag it with the
                                     # new encoding
                                     #
                                     # this is usually a lossless conversion, but
                                     # can sometimes be lossy

A more advanced example:

# encoding: UTF-8
 
# first, tell Ruby that our editor saved the file using the UTF-8 encoding.
# TextMate does this by default. If you lie to Ruby, very strange things
# will happen
 
utf8 = "hellö"
iso_8859_1 = "hellö".encode("ISO-8859-1")
 
# Because we specified an encoding for this file, Strings in here default
# to UTF-8 rather than ASCII. Note that if you didn't specify an encoding
# characters outside of ASCII will be rejected by the parser.
 
utf8 << iso_8859_1
 
# This produces an error, because Ruby does not automatically try to
# transcode Strings from one encoding into another. In practice, this
# should rarely, if ever happen in applications that can rely on
# Unicode; you'll see why shortly
 
utf8 << iso_8859_1.encode("UTF-8")
 
# This works fine, because you first made the two encodings the same

The problems people are really having

The problem of dealing with ISO-8859-1 encoded text and UTF-8 text in the same Ruby is real, and we’ll see soon how it is handled in Ruby. However, the problems people have been having are not of this variety.

If you examine virtually all of the bug reports involving incompatible encoding exceptions, you will find that one of the two encodings is ASCII-8BIT. In Ruby, ASCII-8BIT is the name of the BINARY encoding.

So what is happening is that a library somewhere in the stack is handing back raw bytes rather than encoded bytes. For a long time, the likely perpetrator here was database drivers, which had not been updated to properly encode the data they were getting back from the database.

There are several other potential sources of binary data, which we will discuss in due course. However, it’s important to note that a BINARY encoded String in Ruby 1.9 is the equivalent of a byte[] in Java. It is a type that cannot be reasonably concatenated onto an encoded String. In fact, it is best to think of BINARY encoded Strings as a different class with many methods in common.

In practice, as Ruby libraries continue to be updated, you should rarely ever see BINARY data inside of your application. If you do, it is because the library that handed it to you genuinely does not know the encoding, and if you want to combine it with non-BINARY String, you will need to convert it into an encoded String manually (using force_encoding).

Why this is, in practice, a rare problem

The problem of incompatible encodings is likely to happen in Western applications only when combining ISO-8859-* data with Unicode data.

In practice, most sources of data, without any further work, are already encoded as UTF-8. For instance, the default Rails MySQL connection specifies a UTF-8 client encoding, so even an ISO-8859-1 database will return UTF-8 data.

Many other data sources, such as MongoDB, only support UTF-8 data internally, so their Ruby 1.9-compatible drivers already return UTF-8 encoded data.

Your text editor (TextMate) likely defaults to saving your templates as UTF-8, so the characters in the templates are already encoded in UTF-8.

This is why Ruby 1.8 had the illusion of working. With the exception of some (unfortunately somewhat common) edge-cases, most of your data is already encoded in UTF-8, so simply treating it as BINARY data, and smashing it all together (as Ruby 1.8 does) works fairly reliably.

The only reason why this came tumbling down in Ruby 1.9 is that drivers that should have returned Strings tagged with UTF-8 were returning Strings tagged with BINARY, which Ruby rightly refused to concatenate with UTF-8 Strings. In other words, the vast majority of encoding problems to date are the result of buggy Ruby libraries.

Those libraries, almost entirely, have now been updated. This means that if you use UTF-8 data sources, which you were likely doing by accident already, everything will continue to work as it did in Ruby 1.8.

Digression: force_encoding

When people encounter this problem for the first time, they are often instructed by otherwise well-meaning people to simply call force_encoding("UTF-8") on the offending String.

This will work reliably if the original data is stored in UTF-8, which is often true about the person who made the original suggestion. However, it will mysteriously fail to work (resulting in “�” characters appearing) if the original data is encoded in ISO-8859-1. This can cause major confusion because some people swear up and down that it’s working and others can clearly see that it’s not.

Additionally, since ISO-8859-1 and UTF-8 are both compatible with ASCII, if the characters being force_encoded are ASCII characters, everything will appear to work until a non-ASCII character is entered one day. This further complicates efforts of members of the community to identify and help resolve issues if they are not fluent in the general issues surrounding encodings.

I’d note that this particular issue (BINARY data entering the system that is actually ISO-8859-1) would cause similar problems in Java and Python, which would either silently assume Unicode, or present a byte[], forcing you to force_encoding it into something like UTF-8.

Where it doesn’t work

Unfortunately, there are a few sources of data that are common in Rails applications that are not already encoded in UTF-8.

In order to identify these cases, we will need to identify the boundary between a Rails application and the outside world. Let’s look at a common web request.

First, the user goes to a URL. That URL is probably encoded in ASCII, but can also contain Unicode characters. The encoding for this part of the request (the URI) is not provided by the browser, but it appears safe to assume that it’s UTF-8 (which is a superset of ASCII). I have tested in various versions of Firefox, Safari, Chrome, and Internet Explorer and it seems reliable. I personally thank the Encoding gods for that.

Next, the request goes through the Rack stack, and makes its way into the Rails application. If all has gone well, the Rails application will see the parameters and other bits of the URI exposed through the request object encoded as UTF-8. At the moment (and after this post, it will probably be true for mere days), Rack actually returns BINARY Strings for these elements.

At the moment, Ruby allows BINARY Strings that contain only ASCII characters to be concatenated with any ASCII-compatible encoding (such as UTF-8). I believe this is a mistake, because it will make scenarios such as the current state of Rack work in all tested cases, and then mysteriously cause errors when the user enters a UTF-8 character in the URI. I have already reported this issue and it should be fixed in Ruby. Thankfully, this issue only relates to libraries that are mistakenly returning
BINARY data, so we can cut this off at the pass by fixing Rack to return UTF-8 data here.

Next, that data will be used to make a request of the data store. Because we are likely using a UTF-8 encoded data-store, once the Rack issue is resolved, the request will go through without incident. If we were using an ISO-8859-1 data store (possible, but unlikely), this could pose issues. For instance, we could be looking up a story by a non-ASCII identifier that the database would not find because the request String is encoded in UTF-8.

Next, the data store returns the contents. Again, you are likely using using a UTF-8 data store (things like CouchDB and MongoDB return Strings as UTF-8). Your template is likely encoded in UTF-8 (and Rails actually makes the assumption that templates without any encoding specified are UTF-8), so the String from your database should merge with your template without incident.

However, there is another potential problem here. If your data source does not return UTF-8 data, Ruby will refuse to concatenate the Strings, giving you an incompatible encoding error (which will report UTF-8 as incompatible with, for instance, ISO-8859-1). In all of the encoding-related bug reports I’ve seen, I’ve only ever seen reports of BINARY data causing this problem, again, likely because your data source actually is UTF-8.

Next, you send the data back to the browser. Rails defaults to specifying a UTF-8 character set, so the browser should correctly interpret the String, if it got this far. Note that in Ruby 1.8, if you had received data as ISO-8859-1 and stored it in an ISO-8859-1 database, your users would now see “�”, because the browser cannot identify a valid Unicode character for the bytes that came back from the database.

In Ruby 1.9, this scenario (but not the much more common scenario where the database returns content as UTF-8, which is common because Rails specifies a UTF-8 client encoding in the default database.yml), you would receive an error rather than sending corrupted data to the client.

If your page included a form, we now have another potential avenue for problems. This is especially insidious because browsers allow the user to change the “document’s character set”, and users routinely fiddle with that setting to “fix” pages that are actually encoded in ISO-8859-1, but are specifying UTF-8 as the character set.

Unfortunately, while browsers generally use the document’s character set for POSTed form data, this is both not reliable and possible for the user to manually change. To add insult to injury, the browsers with the largest problems in this area do not send a Content-Type header with the correct charset to let the server know the character set of the POSTed data.

Newer standards specify an attribute accept-charset that page authors can add to forms to tell the client what character set to send the POSTed data as, but again, the browsers with the largest issues here are also the ones with issues in implementing accept-charset properly.

The most common scenario where you can see this issue is when the user pastes in content from Microsoft Word, and it makes it into the database and back out again as gibberish.

After a lot of research, I have discovered several hacks that, together, should completely solve this problem. I am still testing the solution, but I believe we should be able to completely solve this problem in Rails. By Rails 3.0 final, Rails application should be able to reliably assume that POSTed form data comes in as UTF-8.

Moving that data to the server presents another potential encoding problem, but again, if we can rely on the database to be using UTF-8 as the client (or internal) encoding, and the solution for POSTed form data pans out, the data should smoothly get into the database as UTF-8.

But what if we still do have non-UTF-8 data

Even with all of this, it is still possible that some non-BINARY data sneaks over the boundary and into our Rails application from a non-UTF-8 source.

For this scenario, Ruby 1.9 provides an option called Encoding.default_internal, which allows the user to specify an preferred encoding for Strings. Ruby itself and Ruby’s standard libraries respect this option, so even if, for instance, it opens some IO encoded in ISO-8859-1, it will give the data to the Ruby program transcoded to the preferred encoding.

Libraries, such as database drivers, should also support this option, which means that even if the database is somehow set up to receive UTF-8 String, the driver should convert those String transparently to the preferred encoding before handing it to the program.

Rails can take advantage of this by setting the default_internal to UTF-8, which will then ensure that String from non-UTF-8 sources still make their way into Rails encoded as UTF-8.

Since I started asking libraries to honor this option a week ago, do_sqlite, do_mysql, do_postgres, Nokogiri, Psych (the new YAML parser in Ruby 1.9), sqlite3, and the MongoDB driver have all added support for this option. The fix should be applied to the MySQL driver shortly, and I am still waiting on a response from the pg driver maintainer.

In short, by the time 1.9.2-final ships, I don’t see any reason why all libraries in use don’t honor this setting.

I’d also add that MongoDB and Nokogiri already return only UTF-8 data, so supporting this option was primarily a matter of correctness. If a driver already deals entirely in UTF-8, it will work transparently with Rails because Rails deals only in UTF-8.

That said, we plan to robustly be able to support scenarios where UTF-8 cannot be used in this way (because encoding are in use that cannot be transparently encoded at the boundary without data loss), so proper support for default_internal will be essential in the long-term.

TL;DR

The vast majority of encoding bugs to date have resulted from outdated drivers that returned BINARY data instead of Strings with proper encoding tags.

The pipeline that brings Strings in and out of Rails is reasonably well-understood, and simply by using UTF-8 libraries for each part of that pipeline, Ruby 1.9 will transparently work.

If you accidentally use non-UTF-8 sources in the pipeline, Ruby 1.9 will throw an error, an improvement over the Ruby 1.8 behavior of simply sending corrupted data to the client.

For this scenario, Ruby 1.9 allows you to specify a preferred encoding, which instructs the non-UTF-8 source to convert Strings in other encodings to UTF-8.

By default, Rails will set this option to UTF-8, which means that you should not see ISO-8859-1 Strings in your Rails application.

By the time Ruby 1.9 is released in a few months, this should be a reality, and your experience dealing with Ruby 1.9 String should be superior to the 1.8 experience, because it should generally work, but libraries will have properly considered encoding issues. This means that serving misencoded data should be basically impossible.

TL;DR the TL;DR

When using Rails 3.0 with Ruby 1.9.2-final, you will generally not have to care about encodings.

Postscript

With all that said, there can be scenarios where you receive BINARY data from a source. This can happen in any language that handles encodings more transparently than Ruby, such as Java and Python.

This is because it is possible for a library to receive BINARY data and not have the necessary metadata to tag it with an encoding.

In this case, you will either need to determine the encoding yourself or treat it as raw BINARY data, and not a String. The reason this scenario is rare is that if there is a way that you can determine the encoding (such as by looking at provided with the bytes), the original library can do the same.

If you get into a scenario where you know the encoding, but it is not machine available, you will want to do something like:

data = API.get("data")
 
data.encoding #=> ASCII-8BIT # alias for BINARY
 
data.force_encoding("SHIFT-JIS").encode!
 
# This first tags the data with the encoding that
# you know it is, and then re-encodes it to
# the default_internal encoding, if one was
# specified

Posted in Ruby | 19 Comments »

My Common Git Workflow

Thursday, May 13th, 2010

A recent post that was highly ranked on Hacker News complained about common git workflows causing him serious pain. While I won’t get into the merit of his user experience complaints, I do want to talk about his specific use-case and how I personally work with it in git.

Best I can tell, Mike Taylor (the guy in the post) either tried to figure out a standard git workflow on his own, or he followed poor instructions that tried to bootstrap someone from an svn background while intentionally leaving out important information. In any event, I’ll step through my personal workflow for his scenario, contrasting with subversion as I go.

Cloning the Repository

The very first step when working with a repository is to clone it. In subversion, this is accomplished via svn checkout svn://url/to/repo/trunk. This retrieves the most recent revision of the trunk branch of the repository.

In git, this is accomplished via git clone git://url/to/repo (the http protocol is also possible). This retrieves the entire repository, including other branches and tags.

Making the Change

In both git and subversion, you make the change using a normal text editor.

After Making the Change

In git, you make a local commit, marking the difference between the most recent pulled version (master) and the changes you made. In subversion, the normal workflow does not involve making a change, but apparently some people make manual diffs in order to have a local copy of the changes before updating from the remote. Here’s an example comment from the Hacker News post:

I’ll tell you what happens when I use svn and there’s been an upstream change: I never update my local tree with local modifications. Instead, I extract all my local changes into a diff, then I update my local tree, and then I merge my diff back into the updated tree and commit.

When I need three-way merging, which isn’t often – usually patch can resync simple things like line offsets – it’s handled by a file comparison tool. I have a simple script which handles this

My personal process for making the commit in git almost always involves the gitx GUI, which lets me see the changes for each individual file, select the files (or chunks in the files) to commit, and then commit the whole thing. I sometimes break up the changes into several granular commits, if appropriate.

Updating from the remote

Now that we have our local changes, the next step is to update from the remote. In subversion, you would run svn up. Here, subversion will apply a merge strategy to attempt to merge the remote changes with the local ones that you made. If a merge was unsuccessful, subversion will tell you that a conflict has occurred. If you did not manually save off a diff file, there is no way to get back to the status from before you made the change.

In git, you would run git pull. By default, git applies the “recursive” strategy, which tries to merge your current files with the remote files at the most recent revision. As with subversion, this can result in a conflict. You can also pass the --rebase flag to pull, which is how I usually work. This tells git to stash away your commits, pull the remote changes, and then reapply your changes on top one at a time.

If you use --rebase, you may get a conflict for each of your local commits, which is usually easier to handle than a bunch of conflicts all at once.

I definitely recommend using --rebase which also provides instructions for dealing with conflicts as they arise.

In either case, in my experience, git’s merging capabilities are more advanced than subversion’s. This will result in many fewer cases where conflicts occur.

Resolving Conflicts

From here on, I am assuming you followed my advice and used git pull --rebase.

If a conflict has occurred, you will find that if you run git status, all of the non-conflicting files are already listed as “staged”, while the conflicting files are listed outside the staging area. This means that the non-conflicting files are already considered “added” to the current commit.

To resolve the conflicts, fix up the files listed outside the staging area and git add them. Again, I normally use gitx to move the resolved files into the staging area.

Once you have resolved the conflict, run git rebase --continue. This tells git to use the fixed up changes you just made instead of the original commit it was trying to put on top of the changes you got from the remote.

In subversion, if you got a conflict, subversion will create three files for you: file.mine, file.rOLD, and file.rNEW. You are responsible for fixing up the conflicts and getting back a working file. Once you are done, you run svn resolved.

NOTE: If you had not used git pull --rebase, but instead did raw git pull, you would fix up the files, add the files using git add or gitx, and the run git commit to seal the deal

Yikes! Something went wrong!

In git, if something goes wrong, you just run git reset --hard, which will bring you back to your last local commit.

In subversion, it’s not always possible unless you manually stored off a diff before you started.

Pushing

Now that you’re in sync with the remote server, you push your changes. In git, you run git push. In subversion, you run svn commit.

One Glossed-Over Difference

Subversion allows you to commit changes even if you haven’t svn uped and there have been changes to the remote, as long as there are no conflicts between your local files and the remote files.

Git never allows you to push changes to the remote if there have been remote changes. I personally prefer the git behavior, but I could see why someone might prefer the subversion behavior. However, I glossed over this difference because every subversion reference I’ve found advises running svn up before a commit, and I personally always did that in my years using subversion.

Comparison

Here’s a workflow comparison between git and subversion:

Operation	git	svn
Clone a repository	`git clone git://github.com/rails/rails.git`	`svn checkout http://dev.rubyonrails.org/svn/rails/trunk`
Preparing changes	`git commit` (using gitx)	nothing or create a manual diff
Update from the remote	`git pull --rebase`	`svn up`
Resolving conflicts	`git add` then `git rebase --continue`	`svn resolve`
Resolving conflicts without –rebase	`git add` then `git commit`	N/A
Yikes! Rolling back	git reset –hard	`svn up -rOLD` then apply diff (only if you manually made a diff first)
Pushing	git push	svn commit

Note that I am not attempting to provide an exhaustive guide to git here; there are many more git features that are quite useful. Additionally, I personally do a lot of local branching, and prefer to be able to think about git in terms of cheap branches, but the original poster explicitly said that he’d rather not. As a result, I didn’t address that here.

I also don’t believe that thinking of git in terms of subversion is a good idea. That said, the point of this post (and the point of the original poster) is that there are a set of high-level version control operations that you’d expect git to be able to handle in simple cases without a lot of fuss.

Posted in Other | 35 Comments »

The How and Why of Bundler Groups

Sunday, May 9th, 2010

Since version 0.9, Bundler has had a feature called “groups”. The purpose of this feature is to allow you to specify groups of dependencies which may be used in certain situations, but not in others.

For instance, you may use ActiveMerchant only in production. In this case, you could say:

group :production do
  gem "activemerchant"
end

Specifying groups allows you to do two things. First, you can install the gems in your Gemfile, minus specific groups. For instance, Rails puts mysql and pg in a database group so that if you’re just working on ActionPack, you can bundle install --without db and run the ActionPack tests without having to worry about getting the gems installed.

Second, you can list specific groups to autorequire using Bundler.require. By default, Bundler.require requires all the gems in the default group (which is all the gems that have no explicit group). You can also say Bundler.require(:default, :another_group) to require specific groups.

Note the difference between these operations: bundle install is opt-out, while Bundler.require is opt-in. This is because the common usage of groups is to specify gems for different environments (such as development, test and production) and you shouldn’t need to specify that you want the “development” and “test” gems just to get up and running. On the other hand, you don’t want your test dependencies loaded in development or production.

It is also worth noting that all gems that you installed (i.e. not the ones that you excluded at install time with --without) will be available to require. This has no effect unless you actually require them. This means that in development mode, if you explicitly require rspec, it will work.

Rails 3 defaults to mapping groups to environment names, and explicitly autorequiring the implicit default group and the group named the same as the current environment. For example, in development mode, Rails will require the default group and the development group. The code that does this is in your application.rb:

Bundler.require(:default, Rails.env) if defined?(Bundler)

Consistency

In order to ensure consistency across all environments, bundler resolves the dependencies of your application using the gems listed in all groups, even if you specify --without. This means that while you can skip installing the gems listed in the production group by saying --without production, bundler will still download and examine the gems in order to properly resolve all dependencies.

As a result, the dependencies you install in development mode and test with will be compatible with the gems in other environments. In essence, this policy ensures that if your tests pass and run in development, your app will not fail to run in production because the dependencies resolved differently.

Multiple Inconsistent Configurations

Sometimes, especially when developing gems for wider use, you want to test your code against multiple incompatible configurations. At first glance, you might think that you could use groups for this case, but as described above, groups are designed for cases where all of the gems are compatible, but you don’t always want to have to install them in all situations.

Instead, use multiple Gemfiles, one for each incompatible configuration. When installing, do bundle install --gemfile Gemfile.rails2. This will tell Bundler to use Gemfile.rails2 rather than the default Gemfile. As in all cases in Bundler, you can also specify this option globally with an environment variable (BUNDLE_GEMFILE).

Posted in Other | 10 Comments »

Ruby 1.9 Encodings: A Primer and the Solution for Rails

Wednesday, May 5th, 2010

UPDATE: The DataObjects drivers, which are used in DataMapper, are now updated to honor default_internal. Let’s keep this moving.

Since Ruby 1.9 announced support for encodings, there has been a flurry of activity to make existing libraries encoding aware, and a tornado of confusion as users of Ruby and Rails have tried to make sense of it.

In this post, I will lay out the most common problems people have had, and what we can do as a community to put these issues to bed in time for Ruby 1.9.2 final.

A Quick Tour

I’m going to simplify some of this, but the broad strokes are essentially correct.

Before we begin, many of you are probably wonder what exactly an “encoding” is. For me, getting a handle on this was an important part of helping me understand the possible solution space.

On disk, all Strings are stored as a sequence of bytes. An encoding simply specifies how to take those bytes and convert them into “codepoints”. In some languages, such as English, a “codepoint” is exactly equivalent to “a character”. In most other languages, there is not a one-to-one correspondence. For example, a German codepoint might specify that the next codepoint should get an ümlaut.

The list of English characters represented by the first seven bits of ASCII (characters 0 through 127 in “ASCII-7″) have the same representation in many (but not all) encodings. This means that if you only use English characters, the on-disk representation of the characters will often be exactly the same regardless of the source encoding.

However, once you start to use other characters, the bytes on disk mean different things in different encodings. Have you ever seen a page on the Internet filled with something like “FÃ¼hrer”? That is the consequence of the bytes of “Führer” stored as UTF-8 being interpreted as Latin-1.

You can trivially see this problem using Ruby 1.9′s encoding support by running this program:

# encoding: UTF-8
 
puts "hello ümlaut".force_encoding("ISO-8859-1").encode("UTF-8")
 
# Output
# hello Ã¼mlat

First, we create a String (“hello ümlaut”) in the UTF-8 encoding. Next, we tell Ruby that the String is actually Latin-1. It’s not, so an attempt to read the characters will interpret the raw bytes of the “ü” as though they were Latin-1 bytes. We ask Ruby to give us that interpretation of the data in UTF-8 via encode and print it out.

We can see that while the bytes for “hello ” and “mlat” were identical in both UTF-8 and Latin-1, the bytes for “ü” in UTF-8 mean “Ã¼” in Latin-1.

Note that while force_encoding simply tags the String with a different encoding, encode converts the bytes of one encoding into the equivalent bytes of the second. As a result, while force_encoding should almost never be used unless you know for sure that the bytes actually represent the characters you want in the target encoding, encode is relatively safe to use to convert a String into the encoding you want.

You’ve probably also seen the reverse problem, where bytes encoded in Latin-1 ended up inside a page encoded in UTF-8.

# encoding: ISO-8859-1
 
puts "hello ümlaut".force_encoding("UTF-8")
 
# Output
# hello ?mlat

Here, the sequence of bytes that represents an “ü” in Latin-1 could not be recognized in UTF-8, so they were replaced with a “?”. Note that puts will always simply write out the bytes to your terminal, and the terminal’s encoding will determine how they are interpreted. The examples in this post are all outputted to a terminal using UTF-8 encoding.

As you can imagine, this presents quite the issue when concatenating two Strings of different encodings. Simply smashing together the raw bytes of the two Strings can result in output that is incomprehensible in either encoding. To make matters worse, it’s not always possible to represent all of the characters in one encoding in another. For instance, the characters of the Emoji encoding cannot be represented in the ISO-8859-1 encoding (or even in a standardized way onto the UTF-8 encoding).

As a result, when you attempt to concatenate two Strings of different encodings in Ruby 1.9, Ruby displays an error.

# encoding: UTF-8
 
puts "hello ümlaut".encode("ISO-8859-1") + "hello ümlaut"
 
# Output
# incompatible character encodings: ISO-8859-1 and UTF-8 (Encoding::CompatibilityError)

Because it’s extremely tricky for Ruby to be sure that it can make a lossless conversion from one encoding to another (Ruby supports almost 100 different encodings), the Ruby core team has decided to raise an exception if two Strings in different encodings are concatenated together.

There is one exception to this rule. If the bytes in one of the two Strings are all under 127 (and therefore valid characters in ASCII-7), and both encodings are compatible with ASCII-7 (meaning that the bytes of ASCII-7 represent exactly the same characters in the other encoding), Ruby will make the conversion without complaining.

# encoding: UTF-8
 
puts "hello umlat".encode("ISO-8859-1") + "hello ümlaut"
 
# Output
# hello umlathello ümlaut

Since Ruby does not allow characters outside of the ASCII-7 range in source files without a declared encoding, this exception eliminates a large number of potential problems that Ruby’s strict concatenation rules might have introduced.

Binary Strings

By default, Strings with no encoding in Ruby are tagged with the ASCII-8BIT encoding, which is an alias for BINARY. Essentially, this is an encoding that simply means “raw bytes here”.

In general, code in Rails applications should not encounter BINARY strings, except for Strings created in source files without encodings. However, since these Strings will virtually always fall under the ASCII-7 exception, Ruby programmers should never have to deal with incompatible encoding exceptions where one of the two encodings is ASCII-8BIT (i.e. BINARY).

That said, almost all of the encoding problems reported by users in the Rails bug tracker involved ASCII-8BIT Strings. How did this happen?

There are two reasons for this.

The first reason is that early on, database drivers generally didn’t properly tag Strings they retrieved from the database with the proper encoding. This involves a manual mapping from the database’s encoding names to Ruby’s encoding names. As a result, it was extremely common from database drivers to return Strings with characters outside of the ASCII-7 range (because the original content was encoded in the database as UTF-8 or ISO-8859-1/Latin-1).

When attempting to concatenate that content onto another UTF-8 string (such as the buffer in an ERB template), Ruby would raise an incompatible encoding exception.

# encoding: UTF-8
 
puts "hello ümlaut" + "hello ümlaut".force_encoding("BINARY")
 
# Output
# incompatible character encodings: UTF-8 and ASCII-8BIT (Encoding::CompatibilityError)

This is essentially identical to the scenario many people encountered. A UTF-8 String was presented to Ruby as a BINARY String, since the database driver didn’t tag it. When attempting to concatenate it onto UTF-8, Ruby had no way to do so reliably, so it raised an exception.

One reason that many people didn’t encounter this problem was that either the contents of the template or the text from the database were entirely in the ASCII-7 subset of their character set. As a result, Ruby would not complain. This is deceptive, because if they made a small change to their template, or if a user happened to enter non-ASCII-7 data (for instance, they got their first user named José), they would suddenly start seeing an incompatible encoding exception.

When people see this incompatible encoding exception, one common reaction is to call force_encoding("UTF-8") on the BINARY data. This will work great for Strings whose bytes actually are encoded in UTF-8. However, if people whose Strings were encoded in ISO-8859-5 (Russian) followed this instruction, they would end up with scrambled output.

Additionally, it’s impossible to simply encode the data, since Ruby doesn’t actually know the source encoding. In essence, a crucial piece of information has been lost at the database driver level.

Unfortunately, this means that well-meaning people who have solved their problem by force_encoding their Strings to UTF-8 (because the bytes actually did represent UTF-8 characters) become baffled when their solution doesn’t work for someone working on a Russian website.

Thankfully, this situation is now mostly solved. There are updates for all database drivers that map the encodings from the database to a Ruby encoding, which means that UTF-8 text from the database will be UTF-8 Strings in Ruby, and Latin-1 text from the database will be ISO-8859-1 Strings in Ruby.

Unfortunately, there is a second large source of BINARY Strings in Ruby. Specifically, data received from the web in the form of URL encoded POST bodies often do not specify the content-type of the content sent from forms.

In many cases, browsers send POST bodies in the encoding of the original document, but not always. In addition, some browsers say that they’re sending content as ISO-8859-1 but actually send it in Windows-1251. There is a long thread on the Rack tracker about this, but the bottom line is that it’s extremely difficult to determine the encoding of a POST body sent from the web.

As a result, Rack handlers send the raw bytes through as BINARY (which is reasonable, since handlers shouldn’t be in the business of trying to wade through this problem) and no middleware exists (yet) to properly tag the String with the correct encoding.

This means that if the stars align, the raw bytes are UTF-8, end up in a UTF-8 database, and end up coming back out again tagged as UTF-8. If the stars do not align, the text might actually be encoded in ISO-8859-1, get put into a UTF-8 database, and come out tagged as UTF-8 (and we know what happens when ISO-8859-1 data is mistakenly tagged as UTF-8).

In this case, because the ISO-8859-1 data is improperly tagged as UTF-8, Ruby happily concatenates it with other UTF-8 Strings, and hilarity ensues.

Because English characters have the same byte representation in all commonly used encodings, this problem is not as common as you might imagine. Unfortunately, this simply means that people who do encounter it are baffled and find it hard to get help. Additionally, this problem doesn’t manifest itself as a hard error. it can go unnoticed and dismissed as a minor annoyance if the number of non-ASCII-7 characters are low.

In order to properly solve this problem for Ruby 1.9, we need a very good heuristic for properly determining the encoding of web-sent POST bodies. There are some promising avenues that will get it right 99.9% of the time, and we need to package them into up a middleware that will tag Strings correctly.

Incompatible Encodings

If you’ve been paying attention, you’ve probably noticed that while the database drivers have solved one problem, they actually introduced another one.

Imagine that you’re using a MySQL database encoded in ISO-8859-1 (or ISO-8859-5, popular for Russian applications, or any other non-UTF-8 encoding). Now that the String coming back from the database is properly tagged as ISO-8859-1, Ruby will refuse to concatenate it onto the ERB buffer (which is encoded in UTF-8). Even if we solved this problem for ERB, it could be trivially reintroduced in other parts of the application through regular concatenation (+, concat, or even String interpolation).

Again, this problem is somewhat mitigated due to the ASCII-7 subset exception, which means that as long as one of the two incompatible Strings uses only English characters, users won’t see any problems. Again, because this “solution” means that the Ruby developer in question still may not understand encodings, this simply defers the problem to some uncertain point in the future when they either add a non-ASCII-7 character to their template or the user submits a non-ASCII-7 String.

The Solution

If you got this far, you’re probably thinking “Holy shit this encoding stuff is crazy. I don’t want to have to know any of this! I just want to write my web app!”

And you’d be correct.

Other languages, such as Java and Python, solve this problem by encodeing every String that enters the language as UTF-8 (or UTF-16). Theoretically, it is possible to represent the characters of every encoding in UTF-8. By doing this, programmers only ever deal with one kind of String, and concatenation happens between UTF-8 Strings.

However, this solution does not work very well for the Japanese community. For a variety of complicated reasons, Japanese encoding, such as SHIFT-JIS, are not considered to losslessly encode into UTF-8. As a result, Ruby has a policy of not attempting to simply encode any inbound String into UTF-8.

This decision is debatable, but the fact is that if Ruby transparently transcoded all content into UTF-8, a large portion of the Ruby community would see invisible lossy changes to their content. That part of the community is willing to put up with incompatible encoding exceptions because properly handling the encodings they regularly deal with is a somewhat manual process.

On the other hand, many Rails applications work mostly with encodings that trivially encode to UTF-8 (such as UTF-8 itself, ASCII, and the ISO-8859-1 family). For this rather large part of the community, having to manually encode Strings to solve incompatible encoding problem feels like a burden that belongs on the machine has been inappropriately shifted onto Rails application developers.

But there is a solution.

By default, Ruby should continue to support Strings of many different encodings, and raise exceptions liberally when a developer attempts to concatenate Strings of different encodings. This would satisfy those with encoding concerns that require manual resolution.

Additionally, you would be able to set a preferred encoding. This would inform drivers at the boundary (such as database drivers) that you would like them to convert any Strings that they tag with an encoding to your preferred encoding immediately. By default, Rails would set this to UTF-8, so Strings that you get back from the database or other external source would always be in UTF-8.

If a String at the boundary could not be converted (for instance, if you set ISO-8859-1 as the preferred encoding, this would happen a lot), you would get an exception as soon as that String entered the system.

In practice, almost all usage of this setting would be to specify UTF-8 as a preferred encoding. From your perspective, if you were dealing in UTF-8, ISO-8859-* and ASCII (most Western developers), you would never have to care about encodings.

Even better, Ruby already has a mechanism that is mostly designed for this purpose. In Ruby 1.9, setting Encoding.default_internal tells Ruby to encode all Strings crossing the barrier via its IO system into that preferred encoding. All we’d need, then, is for maintainers of database drivers to honor this convention as well.

It doesn’t require any changes to Ruby itself, and places the burden squarely on the few people who already need to deal with encodings (because taking data from outside of Ruby, via C, always already requires a tagging step). I have spoken with Aaron Patterson, who has been working on the SQLite3 driver, and he feels that this change is simple enough for maintainers of drivers dealing with external Strings to make it a viable option. He has already patched SQLite3 to make it respect default_internal.

However you feel about Ruby’s solution to expose String encodings directly in the language, you should agree that since we’re stuck with it for the forseeable future, this solution shifts the burden of dealing with it from the unwashed masses (most of whom have no idea what an encoding is) to a few maintainers of C extensions and libraries that deal in binary data. Getting this right as soon as possible will substantially ease the transition from Ruby 1.8 to Ruby 1.9.

Postscript: What Happened in 1.8!?

When people first move to 1.9 and encounter these shenanigans, they often wonder why everything seemed so simple in Ruby 1.8, and yet seemed to work.

There are a few reasons for this.

First, keep in mind that in Ruby 1.8, Strings are simple sequences of bytes. Ruby String operations just concatenate those byte sequences together without any kind of check. This means that concatenating two UTF-8 Strings together will just work, since the combined byte sequence is still valid UTF-8. As long as the client for the Ruby code (such as the browser) is told that the bytes are encoded in UTF-8, all is well. Rails does this by setting the default charset for all documents to UTF-8.

Second, Ruby 1.8 has a “UTF-8″ mode that makes its regular expression engine treat all Strings as UTF-8. In this mode (which is triggered by setting $KCODE = “UTF-8″), the regular expression engine correctly matches a complete UTF-8 character for /./, for instance. Rails sets this global by default, so if you were using Rails, regular expressions respect unicode characters, not raw bytes.

Third, very little non-English content in the wild is actually encoded in ISO-8859-1. If you were expecting to deal with content that was not English, you would probably set your MySQL database to use a UTF-8 encoding. Since Rails sets UTF-8 as the charset of outbound documents, most browsers will in fact return UTF-8 encoded data.

Fourth, the problems caused when an ISO-8859-1 String is accidentally concatenated into a UTF-8 String are not as jarring as the errors produced by Ruby 1.9. Let’s try a little experiment. First, open up a text editor, create a new file, and save it in the ISO-8859-1 encoding.

$KCODE = "UTF-8"
 
iso_8859_1 = "ümlaut"
 
# the byte representation of ümlaut in unicode
utf8 = "\xC3\xBCmlat"
 
puts iso_8859_1
puts utf8
 
puts iso_8859_1 + utf8
puts utf8 + iso_8859_1
 
# Output
# ?mlat
# ümlaut
# ?mlatümlaut
# ümlaut?mlat

If you somehow get ISO-8859-1 encoded content that uses characters outside of the ASCII-7 range, Ruby doesn’t puke. Instead, it simply replaces the unidentified character with a “?”, which can easily go unnoticed in a mostly English site with a few “José”s thrown into the mix. It could also easily be dismissed as a “weird bug that we don’t have time to figure out right now”.

Finally, Rails itself provides a pure-Ruby UTF-8 library that mops up a lot of the remaining issues. Specifically, it provides an alternate String class that can properly handle operations like split, truncate, index, justify and other operations that need to operate on characters, not bytes. It then uses this library internally in helpers like truncate, transparently avoiding a whole other class of issue.

In short, if you’re dealing mostly with English text, and you get unlucky enough the get ISO-8859-1 input from somewhere, the worst case is that you get a “?” instead of a “é”. If you’re dealing with a lot of non-English text, you’re probably being not using ISO-8859-1 sources. In either case, English (ASCII) text is compatible with UTF-8, and Rails provides solid enough pure-Ruby UTF-8 support to get you most of the rest of the way.

That said, anyone dealing with encodings other than UTF-8 and ISO-8859-1 (Japanese and Russian Rubyists) were definitely not in a good place with Ruby 1.8.

Thanks

I want to personally thank Jay Freeman (aka saurik), who in addition to being a general badass, spent about 15 hours with me patiently explaining these issues and working through the Ruby 1.9 source to help fully understand the tradeoffs available.

Posted in Other | 27 Comments »