Path: csiph.com!eternal-september.org!feeder.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Martin Gregorie <martin@address-in-sig.invalid>
Newsgroups: comp.lang.java.programmer
Subject: Re: Is this the way to work with try with resources
Date: Tue, 8 Aug 2017 12:51:57 -0000 (UTC)
Organization: A noiseless patient Spider
Lines: 101
Message-ID: <omcc5d$uk9$1@dont-email.me>
References: <87poc6wb9n.fsf@munus.decebal.nl>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Tue, 8 Aug 2017 12:51:57 -0000 (UTC)
Injection-Info: mx02.eternal-september.org; posting-host="11a971b7626ac1cfc1a8c8339de20ac6"; logging-data="31369"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/1MkLPmqbbWglugOoo4LpcKcUPOFMRjXs="
User-Agent: Pan/0.141 (Tarzan's Death; 168b179 git.gnome.org/pan2)
Cancel-Lock: sha1:7MYFBnIZx/v3OJ3VwyCeG0JOaL0=
Xref: csiph.com comp.lang.java.programmer:37079

On Tue, 08 Aug 2017 13:26:44 +0200, Cecil Westerhof wrote:

> When using with resources you ‘should’ always work with try with
> resources. At the moment I have something like the following code:
>     private static void statistics() {
>         try (Statement stmt = conn.createStatement()) {
>             String format;
> 
>             try (ResultSet rsTotal =
>             stmt.executeQuery(COUNT_TOTAL_PROV)) {
>                 final int count = rsTotal.getInt("no");
> 
>                 format = String.format("There are %%,%dd %%6s
>                 proverbs.%%n", String.format("%,d", count).length());
>                 System.out.printf(format, count, "total");
>             }
>             try (ResultSet rsUnused =
>             stmt.executeQuery(COUNT_UNUSED_PROV)) {
>                 final int count = rsUnused.getInt("no");
> 
>                 if (count != 0) {
>                     System.out.printf(format, count, "unused");
>                 }
>             }
>             try (ResultSet rsUsed = stmt.executeQuery(COUNT_USED_PROV))
>             {
>                 final int count = rsUsed.getInt("no");
> 
>                 System.out.printf(format, count, "used");
>             }
>         } catch (SQLException e) {
>             System.err.printf("Error in statistics\n");
>             e.printStackTrace();
>             fatalError("");
>         }
>     }
> 
> It feels a bit long-winded. Is this the way to go, or is there a better
> way?
> 
> 
> I thought that I could rewrite it to:
>     private static void statistics() {
>         try (Statement stmt     = conn.createStatement();
>              ResultSet rsTotal  = stmt.executeQuery(COUNT_TOTAL_PROV);
>              ResultSet rsUnused = stmt.executeQuery(COUNT_UNUSED_PROV);
>              ResultSet rsUsed   = stmt.executeQuery(COUNT_USED_PROV)) {
>             int    count;
>             String format;
> 
>             count  = rsTotal.getInt("no");
>             format = String.format("There are %%,%dd %%6s proverbs.%%n",
>             String.format("%,d", count).length());
>             System.out.printf(format, count, "total");
>             count  = rsUnused.getInt("no");
>             if (count != 0) {
>                 System.out.printf(format, count, "unused");
>             }
>             count  = rsUsed.getInt("no");
>             System.out.printf(format, count, "used");
>         } catch (SQLException e) {
>             System.err.printf("Error in statistics\n");
>             e.printStackTrace();
>             fatalError("");
>         }
>     }
> 
In the absence of any other information I prefer the second organisation. 

In general, if you have a set of queries that are related so that if any 
of them fail, the results returned by the others are useless, then its 
sensible to put the lot in a single try...catch block. If the results 
should be treated as independent of whether any of the others fail, then 
they should be in separate try...catch blocks. 

If the queries update the database then all the queries that must either 
all succeed or all fail should be grouped into a commitment unit and this 
should by included in the one try...catch block, since, if there is any 
failure you'll want to roll back the whole commitment unit.

> But there is a problem with that. If I do that, all three result-sets
> have the same query: the last one.
> 
> What is the reason for this? I am using SQLite3.
>
Thats difficult to answer knowing how ACID compliant SQLite3 is and 
whether it is assuming that each query is being run in an automatic 
commitment unit: reuse of "no", presumably a column name, in all three 
statements bothers me a bit and so does the fact that you've executed all 
three queries before asking for any results. As I result of both these 
factors I think you've ended up retrieving the result from the LAST query 
three times. 

It should work as expected if each query is executed immediately before 
calling getInt() to  retrieve and print its result.


-- 
martin@   | Martin Gregorie
gregorie. | Essex, UK
org       |