929356 - [zffos1.1][contacts] -The phone number matching rule for MMS/SMS/Dailing is not fit for Venezuela now.

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: Gaia::Contacts, defect)

Description

[zhangdapeng]

Attachment: logs.zip

User Agent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0; .NET CLR 2.0.50727; .NET CLR 3.0.4506.2152; .NET CLR 3.5.30729; TCO_20131022155504)

Steps to reproduce:

The phone number matching rule for MMS/SMS/Dailing is not fit for Venezuela now. 

For example, here saves a number 01234567891 named IVAN. The SMS/MMS message number from IVAN will be 581234567891 (first number 0 change to 58), 


 



Actual results:

Cause now  matching rule matches 11 number, so the number 581234567891 is not recognized as IVAN. 



Expected results:

It matches 8 number before,and  the Venezuela said 10 number matching is needed.

Comment 1

[shawn ku [:sku]]

Hi Gene:
 Could you please take a look of this bug?

The SIM is Venezuela SIM (MCC: 734), and the MT SMS with sender=[phone] can not be recognized as a saved contact.

Thanks!!
sku

// b2g.js
pref("dom.phonenumber.substringmatching.VE", 7); 

// log
10-21 08:42:50.929 D/RIL_QC_B2G(  112): [0055]< GET_IMSI {734044700283614}
10-21 08:42:51.429 D/RIL_QC_B2G(  112): [0088]< OPERATOR {Telefonica Moviles Venezuela,movistar,73404}

10-21 08:47:18.759 D/RIL_QC_B2G(  112): [UNSL]< UNSOL_RESPONSE_NEW_SMS { SMSC=584240000954, mti=0, udhi=0, sender=[phone], receiver=, pid=0, epid=0, dcs=0, header={length : 0, destinationPort: -1, originatorPort: -1, segmentRef: 0, segmentMaxSeq: 0, segmentSeq: 0, langIndex: 0, langShiftIndex: 0, }, body=Test, data=, messageRef=0, status=0, encoding=0, isMwi=0 Year=2013, mon=9, day=21, hour=8, minute=47, second=16}
10-21 08:47:18.889   112   112 D SMS_QC_B2G: Done notifying DOM of received SMS message: sender [phone], body Test

Comment 2

[shawn ku [:sku]]

By using fake IMSI: 734044700283614 and MCC/MNC: 73404 for camped operator, 
Saved contact: Test/04265129005

[phone] can be recognized as "Test" when MO call or SMS on both v1.1 and v1.2.

Symptom can not be founded.

Comment 3

[zhangdapeng]

the problem was founded after the September 10th version ,it was right before.
you can see the logs.zip,it was the log from  Venezuela corresponding to this problem.

Comment 4

[Gene Lian (I already quit Mozilla)]

Sorry for the delayed response. I have been struggling with some V1.3 features.

Actually, we recently improved some phone number matching logic at Bug 914060. I wonder that fix can also cover this bug.

Hi Bevis, since you're solving related issues at bug 934931, it sounds more efficient for you to take this one as well. Tentatively assign this one to you. Please feel free to have discussion with Ctai or me when you encounter any difficulties. Thanks! :)

Comment 5

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Zhang,

From Comment#2, Comment#3, and the last change date of mozilla-b2g18_v1_1_0 found in our server (https://hg.mozilla.org/releases),
It seems that there is no change after Apr, 02, 2013.

Could you help to narrow the the versions between which this issue is reproducible?
Besides, is it still reproducible in latest version? (According to Comment#2, it seems workable in the ROM we have at the end of Oct.)
If yes, could you help to provide the corresponding version information for us to double check?

Thanks,
Bevis

Comment 6

[zhangdapeng]

it is still reproducible in latest version，
save a number 01234567891 named “test”. 
dial number 11234567891 can not be recognized as  “test”
our September 8th daily build version can recognize ，but can't after September 10th version.
in the September 9th,our version had a upgrade

Comment 7

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Zhang,

Thanks for your information. It seems that the v.1.1 shall be the one in mozilla-b2g18_v1_1_0_hd instead and is still updating recently.

Hi Gregor,

After checking the change log of PhoneNumberUtils.jsm in 1.1.0hd, I found that there was some modification on 09/03 for Bug 905927 to change the number matching rule for the operators in Venezuela.
Would you please kindly help us to clarify if this bug is related to the change?

Thanks & regards,
Bevis Tseng

Comment 8

[Bevis Tseng[:bevis][:btseng](Exited)]

CC

Comment 9

[Bevis Tseng[:bevis][:btseng](Exited)]

CC:Anshul, Michael, because this might need help from Commercial RIL.

Comment 10

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Gregor,

After reviewing the implementation, I found that in PhoneNumberUtils.getCountryName(). It will try to get the mcc from Services.prefs.getCharPref("ril.lastKnownSimMcc");

However, currently, this prefs is set in RadioInterfaceLayer.js and might not be available in commercial RIL.

Need your help to comment if this will impact the feature of number maching in Venezuela Operators.

Regards,
Bevis Tseng

Comment 11

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Zhang,

Can you help to provide more detail information about the your daily build version info on 9/8 and 9/9 in the following example between which this problem is still occurred?

GAIA_REV=355be5eebe5bb80f01d57076742c6e3beb8f8d0d
GECKO_REV=6529acef3ef8c7e966cdefb86b91310941576c01
GAIA_BRANCH=mozillaorg/v1.1.0hd
GECKO_BRANCH=mozillaorg/v1.1.0hd
BuildID=20131108020037

We would like to narrow down which patch in which branch might cause this regression problem.

Thanks & regards,
Bevis Tseng

Comment 12

[zhangdapeng]

hi
We do not know how to get GAIA_REV and GECKO_REV，but updated to your lastest code at that time in  9/9 ，

Comment 13

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Zhang,

You can get these information by the following command inside the gecko folder of your code to get the GECKO_REV:
#cd $B2G_ROOT/gecko
(Where $B2G_ROOT is the path of your B2G folder)
#git branch -a
This help's you to print out the brach of the gecko you use.
#git log
(The 1st commit hash is the revision of your local work directory)
#cat .git/config
(the url specifies the git server you pull from)

Thanks & regards,
Bevis Tseng

Comment 14

[zhangdapeng]

Attachment: git log.txt

does it help？

Comment 15

[Bevis Tseng[:bevis][:btseng](Exited)]

No, it seems more likely to be the internal ZTE revision from the git log and we don't know what the corresponding b2g version you pull from Mozilla.

Can you provide the command of how you get the b2g v1.1 from mozilla?

Bevis

Comment 16

[zhangdapeng]

GAIA_REV=3335e3ba4e8bb6965112514f00e8ba4e312ab540
commit 3335e3ba4e8bb6965112514f00e8ba4e312ab540
Merge: 5b1b385 9df4b6b
Author: David Flanagan <[e-mail]>
Date:   Wed Aug 28 17:43:42 2013 -0700

    Merge pull request #11821 from davidflanagan/bug890427-v1-train
    
    Bug890427 v1 train

GECKO_REV=30cc5958c2a8ff8fc25c7166022e81d4940d185f
commit 30cc5958c2a8ff8fc25c7166022e81d4940d185f
Author: Sasikala Paruchuri <[e-mail]>
Date:   Fri Sep 6 19:13:57 2013 +0800

    Bug 912517 - [Message] Crash occuring in MMS when killing the application while the MMS is sending. r=vicamo, a=leo+

this can see in the "git log.txt"

use git merge

Comment 17

[Bevis Tseng[:bevis][:btseng](Exited)]

Attachment: Sucessful Log of phone number matching in v_1_1_hd

I've attached a log which proves that in the v_1_1_hd ROM, this is workable to match the number when the mcc/mnc of SIM/Network are both 73404.

Here is the branch and gecko revision info:
GECKO_REV=6529acef3ef8c7e966cdefb86b91310941576c01
GECKO_BRANCH=mozillaorg/v1.1.0hd


11-08 16:20:33.773 I/Gecko   ( 3429): -*- PhoneNumberutils: MCC: 734countryName: VE
11-08 16:20:33.783 I/Gecko   ( 3429): -*- ContactDB component: contact:{"properties":{"name":["IVAN  "],"honorificPrefix":[],"givenName":["IVAN "],"additionalName":[""],"familyName":[],"honorificSuffix":[],"nickname":[],"photo":null,"category":[],"tel":[{"type":["mobile"],"value":"01234567891","carrier":""}],"org":[],"jobTitle":[],"note":[],"sex":null,"genderIdentity":null},"search":{"givenName":["ivan "],"familyName":[],"email":[],"category":[],"tel":["01234567891","1234567891","234567891","34567891","4567891","567891","67891","7891","891","91","1"],"exactTel":["01234567891"],"parsedTel":["01234567891","4567891"]}}
11-08 16:21:08.007 I/Gecko   ( 3429): -*- ContactDB component: ContactDB:find val:[phone] by: tel op: match
11-08 16:21:08.007 I/Gecko   ( 3429): -*- ContactDB component: _findWithIndex: [phone] match tel
11-08 16:21:08.017 I/Gecko   ( 3429): -*- ContactDB component: key: tel
11-08 16:21:08.017 I/Gecko   ( 3429): -*- ContactDB component: match
11-08 16:21:08.017 I/Gecko   ( 3429): -*- PhoneNumberutils: normalize([phone]): [phone], true
11-08 16:21:08.017 I/Gecko   ( 3429): -*- ContactDB component: Substring: 4567891
11-08 16:21:08.027 I/Gecko   ( 3429): -*- ContactDB component: Request successful. Record count: 1

Comment 18

[Bevis Tseng[:bevis][:btseng](Exited)]

Hi Zhang,

We can match the number correctly from the log I've attached with the revision info as followed if I fake the MCCMNC as '73404':
Here is the branch and gecko revision info:
GECKO_REV=6529acef3ef8c7e966cdefb86b91310941576c01
GECKO_BRANCH=mozillaorg/v1.1.0hd

Can you use the pure Mozilla v_1_1_hd build instead of ZTE integrated version to verify if the problem  still exists?

Bevis Tseng

Comment 19

[Gregor Wagner [:gwagner]]

(In reply to Bevis Tseng [:bevistseng] ([e-mail]) from comment #10)
> Hi Gregor,
> 
> After reviewing the implementation, I found that in
> PhoneNumberUtils.getCountryName(). It will try to get the mcc from
> Services.prefs.getCharPref("ril.lastKnownSimMcc");
> 
> However, currently, this prefs is set in RadioInterfaceLayer.js and might
> not be available in commercial RIL.
> 
> Need your help to comment if this will impact the feature of number maching
> in Venezuela Operators.
> 

lastKnownSimMcc is a backup when we don't have the current country code from the sim card.
I was just looking at http://mxr.mozilla.org/mozilla-central/source/dom/phonenumberutils/PhoneNumberUtils.jsm#77
For non-moz RIL we rely on lastKnownSimMcc. The perfect solution would be if we could read mcc from the sim card here.

Comment 20

[Francisco Jordano [:arcturus] [:francisco]]

Hei Gregor, what should be the next steps here?

Comment 21

[Gregor Wagner [:gwagner]]

(In reply to Francisco Jordano [:arcturus] [:francisco] from comment #20)
> Hei Gregor, what should be the next steps here?

We have to set ril.lastKnownSimMcc for commercial RIL as well. Is 
http://mxr.mozilla.org/mozilla-central/source/dom/system/gonk/RadioInterfaceLayer.js#2807
only for moz ril?

Comment 22

[BMO Automation]

Firefox OS is not being worked on