[IOTDB-127] Update some execution result pictures in documents

update jdbc doc

[IOTDB-127] Fix documents problems

[IOTDB-145] Update readme.md and readme_zh.md in incubator-iotdb/grafana

[IOTDB-146] Windows should use different mvn command, like using quoting '-Dmaven test.skip=true'

[IOTDB-141]Develop memory estimate tool

* Add TsFile Docs, delete Json schema

improve some code in UnseqResourceMergeReader

[IOTDB-142] Change String literal "UTF-8" into StandardCharsets.UTF_8

add documents to explain how to use IDEA/Eclipse to develop IoTDB

- avoid the creation of TsFileSequenceReader for skipped resources
- avoid the creation of DigestForFilter when the filter is null

Cherry picks from rel/0.8 to master

Fix errors in 1-Deployment.md

Update 0-Content.md

perfect documents

* Refactor the reader package for readability

The former code in the org.apache.iotdb.db.query.reader package is
somewhat diffcult to read.

Through this refactor, two main tasks are completed:

* the three levels of readers are made explicit: chunkRelated <
resoureRelated < seriesReader

* the two interfaces for creating the unsequence readers in
SeriesReaderFactoryImpl are removed. Instead, two classes UnseqResourceMergeReader
and UnseqResourceReaderByTimestamp are used. This change brings two
benefits: a) SeriesReaderFactoryImpl is thus focused purely on the
creation of series readers and acts as a middleman between query
executor and data reader.b) The classes for sequence and unsequence
classes are symmetric finally.

The future work includes further removing duplicated code and writing
better annotations.

* Deprecate the SeriesReaderFactoryImpl

The former code creates the three series related readers (
SeriesReaderWithoutValueFilter, SeriesReaderWithValueFilter and
SeriesReaderByTimestamp) by the help of SeriesReaderFactoryImpl.
However, thanks to the latest refactor of the reader package, it becomes
clear that the creation tasks can be handed over to the constructors
of the three series reader classes respectively. This change brings about
two benefits: 1) the role of the three series readers are more
clear and controllable since now the inputs of their constructors are not
pre-created readers but the raw resouces. 2) the SeriesReaderFactoryImpl
now can be removed or changed to perform other tasks such as code
deduplication.

update

* Add JAVA annotations and improve related tests.

reformat license

* fix the bug in getValueInTimestamp of SeqResourceReaderByTimestamp; further split the subpackage chunkRelated into two subpackages chunkRelated and fileRelated.

* unsealed sequence TsFile's endTimeMap can also be used in constructNextReader of SeqResourceReaderByTimestamp to speed up

* add resource filter step in the resource related reader construction process

* add more annotations for isTsFileNotSatisfied and constructNextReader

(cherry picked from commit 50f2da57)

(cherry picked from commit 76476bb4)

Fixed a bug where all tests in IoTDBAggregationIT fail on a System with Locale different than English.

(cherry picked from commit 1903150f)

(cherry picked from commit dd4ecf28)

(cherry picked from commit f1442f41)